Berthier - Iridescences-The Physical Colors of Insects

Download as pdf or txt
Download as pdf or txt
You are on page 1of 170

Iridescences

Serge Berthier

Iridescences
The Physical Colors of Insects
Serge Berthier
Institut des Nano-Sciences de Paris
UMR CNRS - Universite Pierre et Marie Curie
Universite Denis Diderot
4 Place Jussieu
Paris, 75252
France

Translator
Capucine Lafait
7 Rue Christian Dewet
Paris, 75012
France

Library of Congress Control Number: 2006925252

ISBN-10: 0-387-34119-6 e-ISBN-10: 0-387-34120-X


ISBN-13: 978-0-387-34119-4 e-ISBN-13: 978-0-387-34120-0

Printed on acid-free paper.


C 2007 Springer Science+Business Media, LLC
All rights reserved. This work may not be translated or copied in whole or in part without the written permission of
the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief
excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and
retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter
developed is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identied
as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.

9 8 7 6 5 4 3 2 1

springer.com
About the Author

Serge Berthier teaches physics at the Denis Diderot-Paris 7 University and researches biologic
structures, colors and biomimetism at the Institut des NanoSciences de Paris (Pierre and Marie
Curie-Paris 6 University and CNRS). He teaches solid state optics in the post-graduate degree
Optics and Material and the post-graduate research degree Optics and Photonic for the Denis
Diderot University. He also teaches electromagnetism and laser physics to undergraduates at the
Paris-Jussieu Institut of Technology.

v
Foreword

Here is a book to recommend for various reasons. First of all, let us note that the rst chapter
introduces the reader into a reection on the physical explanation of an observation. The origin
of buttery colors has interested a great physics experimenter, A.A. Michelson, Nobel Prize in
1907, thought it proceeded from selective reection over buttery wings and defended this idea
during his whole life. The book he wrote in 1927 at the age of 75 and which contains all of
the conclusions of his work, ends with a chapter entitled metallic colors of birds and insects.
At the same time, another great physicist, Lord Raleigh, claimed that observed colors were
caused by light interferences. How is it possible to defend two different explanations for similar
observations? Simply because the knowledge of the objects observed was inappropriate. Here is a
simple case which shows how a scientic explanation can evolve with the knowledge of the object
observed.
S. Berthiers book beautifully shows that it is possible to teach physics in relation with biology,
which makes the latter certainly more attracting and lively. One is here offered an optics course
through a treatise on Lepidoptera and Coleopteron colors. This shows that it is possible to teach
physics in a painless way.
Lastly, the conclusion underlines that the study of buttery iridescence entice researchers
to nd ways to apply the same principles to textiles. It also highlights that Butteries and
Cincidelidaes properties could be used to ensure the identication and protection of bank
notes.
As founder of the laboratory in which S. Berthier works, I am very happy and proud to present
this book. Its author has a gift for pedagogy and this book is fascinating at all levels. It should
interest young people, as young as high school seniors, but also every people who likes to learn
and understand nature further. This book succeeds in being concise and clear, but also in making
one marvel. It is also a step forward as concerns pluridisciplinarity. I wish that it is successful, as
it deserves, and I congratulate the author.

Florin Abeles
Professor Emeritus of Pierre et Marie University, Paris.

vi
Acknowledgments

Part of this book is a new, improved, and substantially modied version of the rst book, Buttery
Colors or the Imperative Beauty. I wont mention again all the people who contributed to the rst
book. They will recognize their contribution in the present work that made up Imperative Beauty
and was the basis of Iridescence. I would like to thank them once again, the pupils, students,
and teachers who gave a reason for being to these publications. I thank my colleague and friend
Michel Perreau for his uncompromising correcting of the rst book and making pertinent remarks
about the present book, which were very useful to me.
Other people took over, furthering and widening the scope of the rst exploratory work.
A schoolboy also whose stay was too short: Antoine Baillet. Thank you, my dear Antoine!
Spadework, exalting at rst but also quite laborious, was performed joyfully and contentedly by
Aurelie Tournie and Marius Knoch, students at the Institut Universitaire de technologie (IUT)
Paris-Jussieu, and Mathieu Gueguen at IUT of Annecy. They made a great number of the most
beautiful of the photonic microscopy photos presented here. I thank them for their patience and
congratulate them for succeeding in conciliating scientic rigor and artistic sensitivity.
Postgraduate students focused more on applications. Passing from quality to quantity does
not always prove uplifting. If their conclusions have only a small place in the present book, they
still made the latter possible in terms of logistics. Thanks to Jean Druille and Loic Ledernez,
students of the Optics and Materials Master at the University Denis Diderot in Paris.
We aimed at widening the scope of the study of iridescence beyond physics and butteries
in the present publication, which is beyond our qualications. That is why we relied on the
work of specialists, especially Pr. Henri Descimon, of Marseille University, on pigments. He also
provided us with a certain number of photos illustrating his conclusions. I want to respectfully
thank him.
I would also like to thank my friend Emanuel Fritsch of the Institut des Materiaux de Nantes
for photos of opals, as well as Philippe Lalanne of the Institut dOptique dOrsay, my colleague
of the Optics and Photonic post-graduate degree, for the calculations of electromagnetic elds
diffracted by the Morphos.
These research works were performed at the color and biomimetism group of the Labora-
toire dOptique des SolidesUniversite Pierre et Marie Curie, Paris. I thank all the staff of the
laboratory, beginning with its director, Jacques Lafait, for his unbreakable support. Some people
directly contributed to this work, like Claude Sella, who suggested I should use ionic thinning
on elytrons, thus allowing me to distinguish in relief chitin sticks. These elytrons had rst un-
dergone optical polishing in our crystallography workshop in Michelle Jacquets expert hands,
who I friendlily thank. A special thanks to Eric Charron, who performed spectogoniometric
measurements and established Morphos diffraction maps, as well as to those who patiently and
kindly tried to pacify my relationship with computers and get me out of the incredible messes I
get myself in: Claude Naud and Gerard Vuye.

vii
viii Acknowledgments

Most of the scanning electron microscopy photos were taken by Stephane Borensztein at the
Laboratoire de Physique des Liquides et Electrochimie at the University Pierre and Marie Curie.
Transmission electronic or photonic microscopy sections of scales and alary membranes were
made at the Laboratoire de Biologie Marine at the same university. Thanks to Jean-Pierre Lechair
and Ghislaine Frebourg for welcoming me and for their beautiful work.
I was about to forget my friend and ofcial supplier of butteries and other insects, Claude
Nature. You can go to his shop 6, rue des Chantiers in the 5th in Paris to get a good dose of marvel
and good spirits.
Lastly, authors are unbearable persons. Thanks to my children, Valerian and Juliette, for putting
up with my unavailability and sudden changes of moods with so much kindness and humor.
Annie Fontaine had her share, and in addition, she patiently and rigorouslyand tactfully,
authors being touchy personscorrected proofs. Thank you for the warmth and tenderness you
gave me.
Contents

1 Iridescence ...................................................................................................... 1
About the Book................................................................................................. 5

2 Why Colors...................................................................................................... 7
Variations in Colors ........................................................................................... 9
Colors: Another Classication ............................................................................. 10
Mimicry........................................................................................................... 10

3 Lepidoptera Description and Scales of Observation.............................................. 14


Outlining a Classication ................................................................................... 14
Summary Description of Butteries...................................................................... 15
Description of the Wings .................................................................................... 20

4 Coleoptera Description and Observation Scales ................................................... 40


Coleoptera ....................................................................................................... 40

5 Changing Colors: Structures or Pigments? ........................................................... 49


A Few Basic Experiments ................................................................................... 49
Pigments.......................................................................................................... 49
A Few General Principles ................................................................................... 52
Changing Colors ............................................................................................... 54

6 Physical Colors, Chemical Colors Basics of Solid State Optics ............................... 56


LightMatter Interaction, Polarization, and Optical Index ...................................... 56
Structural Origin of Colors; What Kind of Logic for Structures?................................ 65

7 1-Dimensional Structures: Interferences.............................................................. 68


Theory Recalls .................................................................................................. 68
Type Butteries with Convex Structures ............................................................... 72
Dispersion and Polarization................................................................................ 77
Colored Effects ................................................................................................. 77
Concave Structures: Papilio and Cincidelae .............................................................. 78
Papilionidae: Solid/Gas Structures ........................................................................ 78
Cicindelidae: Solid/Solid Structures....................................................................... 82
Cetonia: Circular Polarization .............................................................................. 82

ix
x Contents

8 2-Dimensional Structures: Interferences and Diffraction ..................................... 86


Theory Recalls................................................................................................. 86
Type Butteries with 2-Dimensional Structures .................................................... 88
Structures of Wings and Scales........................................................................... 92
Dispersion and Polarization .............................................................................. 98
Evolution of the Structures................................................................................ 104
Polarization Effects .......................................................................................... 105

9 3-Dimensional Structures: Crystalline Diffraction .............................................. 112


Theory Recalls................................................................................................. 112
Polarization Effects .......................................................................................... 114
Insects and Photonic Crystals ............................................................................ 115

10 Amorphous Structures: Scattering..................................................................... 117


Theoretical Recalls ........................................................................................... 117
Type Butteries with Scattering Structures........................................................... 123

11 Pigments and Pigmentary Colors ...................................................................... 127


Selective Absorption and Colors of Pigments ....................................................... 127

12 Thermoregulation and Spectral Selectivity......................................................... 135


Imperatives .................................................................................................... 135
Energetic Balance............................................................................................. 136

13 Vision and Colorimetry ................................................................................... 142


Color Perception.............................................................................................. 142
Sources and Illuminants.................................................................................... 143
Colorimetry Notions ........................................................................................ 144
Buttery Vision ............................................................................................... 146

Conclusion........................................................................................................... 151
Subject Index ....................................................................................................... 157
General Index....................................................................................................... 158
1
Iridescence

Solid state physics is traditionally classied species, mostly tropical. Coleoptera and even
among hard sciences, in the literal sense by certain birds can actually surpass them as far as
students and in the gurative sense by the this aspect is concerned. It is iridescence, an im-
Academy. To introduce the study of optical propriate term in French, yet a last allusion to
properties of buttery wings is, at best, mere the Ancients and the messenger of the Olympic
wandering and at worst, provocation. Since I gods, the winged goddess Iris draped in her
have started this research, I have been told so rainbow. From the very end of the 19th cen-
many times, though always politely, that in ac- tury, iridescence started to intrigue scholars:
cordance with the current symbolism of butter- biologists of course, who gave the name of the
ies, I was itting around. Yet, butteries have goddess to a great number of species present-
not always stood for inconsistency, futility and ing the qualitywhich we tried to illustrate in
banter! This view appeared as late as in the 17th this bookbut also physicists. If artists would
century and applies to diurnal butteries only. then stick to a frivolous symbolism, things were
Moths are not luckier though. In his introduc- much less easy on the other side! On one side
tion to devotion life (1608), Saint-Francois de the poet:
Sales, a connoisseur, immortalizes the symbol-
ism of moths: Natre avec le printemps, mourir avec les roses,
Sur laile du zephyr nager dans un ciel pur
As the small buttery perceiving the candlelight goes Balance sur le sein des eurs a peine ecloses,
towards it in a peculiar manner to see if it is as soft as it is Senivrer de parfums, de lumiere et dazur,
beautiful, and hurried by his fantasy, never stops unless Secouant, jeune encore, la poudre de ses ailes,
it gets lost on its rst attempt: so the young heart goes for Senvoler comme un soufe aux voutes eternelles;
voluptuous ames. Voila du papillon le destin enchante!
Il ressemble au desir qui jamais ne se pose
The Ancients had a completely different Et sans se satisfaire, efeurant toute chose
view, which is more relevant to the purpose Retourne enn au ciel chercher la volupte.
of this book: The buttery embodies immor- Lamartine, Nouvelles meditations poetiques
tality of the human being. In Greek, the word
psyche designates both the human soul and To be borne with the spring, and die with roses,
the buttery. Metamorphosis, the mysterious On zephyr wings to swim in a clear sky
and obscure passage from earth to ether, offers Rocked in the hearts of blooming owers,
To get drunk on fragrances, on light and azure,
a simple symbolism to the rst Christians who
Flapping the dust out of its new borne wings,
adopted it. To y like a breath to the eternal skies,
But let us come back to earth for a mo- Here is the enchanted fate of the buttery!
ment, and return temporarily to physics. One It is like desire which never rests,
of the most fascinating aspects of buttery col- And never contented, always eeting
ors, in addition to their innite diversity, is cer- At last goes back to the sky to nd delight.
tainly the metallic and changing aspect of some Lamartine, New poetic meditations

1
2 Chapter 1 Iridescence

Figure 1.2. The buttery cherub by William


Bouguereau: It looks like desire . . .

awarded the Nobel Prize in science (1907). Pres-


ident of the American Association for the Ad-
vancement of Science and soon president of
the National Academy of Sciences, his position
made him somehow haughty:

It is somewhat surprising to nd that the contrary view


[he was a passionate advocate of the theory of se-
lective reection, which will be mentioned later on]
Figure 1.1. Representation of soul holding fruit of is still held by eminent naturalists and it is hoped that
paradise. Roman catacombs at Domitille, second or further evidences here presented may serve to emphasize
third century (after a Charbonneau-Lassay drawing: the distinction between metallic or surface colors and the
The Christ bestiary).

And on the other side, questions: Why these


colors? And how such colors? Then, of course,
things started to take a turn for the worse. The
cacophony of the homo technicus is opposed
to Lamartines almost immortal verses.
Here is what professor R.W. Wood said in
1919 concerning the structural origin of certain
colors, clearly separated from pigmentary col-
ors:
Not only are the existing theories inadequate to explain
the phenomena but these (colors) cannot be explained by
any of the common laws of optics with which we are fa-
miliar. (Phil. Mag., London, 38 (1919) 98)

And we must admit that the controversy Figure 1.3. Scales of the dorsal side of Morpho
that opposed two physicists who were among godarti. Picture obtained by transmission photon
microscopy. Each scale is about 200 m long by
the greatest of the time could puzzle people 50 m wide. One can distinguish the main longitu-
and make them cautious. Let us start with dinal striae network and, in places, the transversal
Michelson himself, the rst American to be counter-striae network.
Iridescence 3

remaining classes of colors: interferences, diffraction . . . colors rules out any kind of visualization. The
(Phil. Mag. 21 (1911) 554) problem can only be addressed through optical
measurements of the whole wing as it enables
From the other side of the Atlantic Ocean, an-
us to track down the nature and structure of
other famous name, Lord Raleigh (4th baron),
its components. This faction is mainly com-
president of the Royal Institution and professor
posed of physicists, like B. Walter, and Bie-
at the Imperial College of London, and advo-
dermann, a biologist who will later side with
cate of the interferential approach, replied:
Raleigh. Michelson based his theory on mea-
It is singular that the explanation of some of the most surements of polarization of light reected by
striking and beautiful of optical phenomena should be still the wing under various angles of incidence (el-
matters of controversy! (Phil. Mag. 37 (1919) 98) lipsometric measurements, as one would call
them nowadays) very similar to those obtained
In order to understand how great physicists
on thin lms of aniline, or magenta. This was a
can have quarreled over such a subject, let us re-
forceful argument, which has yet to be turned
member that at the turn of the 20th century, the
back.
only means of observation available was tradi-
In the other faction were the defenders of
tional photonic microscopy, which was indeed
structural origin. In this faction, Lord Raleigh
unable to resolve structures capable of creat-
and numerous biologists attempted to make
ing such colors, whatever they were. The im-
out this structure. It is to them that we owe ex-
age it could give of a scale resembled the one
tremely precise microscopic observations and
depicted above, and originated many a mis-
outstanding drawings of the scales of a great
take. One can clearly distinguish longitudinal
number of butteries and of sections that
striaeand even a square pattern sometimes
they did thanks to rather rudimentary mi-
which immediately reminds one of a grating.
crotomes. The weightier argument in favor
Hence the hypotheses of diffraction that were
of the thin lm hypothesis as defended by
made at a certain time. Two types of experi-
Lord Raleighwhich we will fully illustrate
mental approaches were then considered and
in this bookis the alteration or even the to-
two factions emerged opposing the different
tal disappearance of colors when scales are
protagonists of the controversy.
immersed in a liquid with an adequate opti-
One faction was headed by Michelson and
cal index. This implies that the structure re-
included those who think that colors result
sponsible for colors remains open, enabling
from absorption and selective reection of a
the liquid to penetrate it. This was not eas-
metallic type. In almost all of their papers, they
ily conceivable, but it excluded the hypothe-
refer to thin lms of aniline or cianinne The
sis of selective reectionwhich, on the con-
rst consequence of which is that it is not nec-
trary, would imply even brighter colors in this
essary to carry a minute observation of scales,
case. It is quite hard to understand how physi-
since the atomic or molecular origin of the
cists like Michelson and Walter could have ne-
glected this question. Another physical source
of colors, among which skystudied by Lord
Raleigh the fatheris a familiar example is the
selective scattering by small particles in numer-
ous living beings (feathers of certain birds,
muzzle of mandrills, human iris, and so on).
Tyndall observed this as early as the late 19th
century (1873), and so did Madoul in 1903.
When these particles are small enough, they
Figure 1.4. Structures and pigments share the visi- only scatter the shortest wavelengthspurple
ble spectrum for the production of colors among but-
teries. Green, in the middle of the spectrum, often and bluethe longest being transmitted and
results from combined effects. This gives the process eventually absorbed by deep layers. This is
a large exibility and allows a cheap adaptability. the case of many Lycaenidae, one of the rare
4 Chapter 1 Iridescence

examples of blue butteries living under our (the majority of semi-precious stones) or inter
tempered climate. When larger, the particles or intra-band transitions of crystalline material
scatter longer wavelengths and we obtain the (for example, metals and semiconductors). If
white of the Pieridae. effects like camouage are obviously useful,
Those physical phenomena generally cause we can wonder why a buttery develops very
cold colors, or high energy colors, ranging loud motifs or colors. Butteries belong to the
from near ultra-violet to blue-green, while pig- class of insects, which entails two important
ments usually create hot colors, from red to consequences. First, because of their position
yellow. Indeed, blue pigments are rare in the in the food chain, they undergo great selective
animal world: there is a carotenod in the shells pressure. Second, thanks to their short life
of some crustaceans, like lobster, while haemo- span and high reproductive rate, they can
cyanin is quite widespread among mollusks, rapidly evolve. Colors are, of course, only
crustaceans and arachnids. As we will see in the a small element of the gear that butteries
chapter on pigments, the scarcity of blue pig- develop throughout their ght for life. One
ments in the animal world can easily be under- still remains fascinated by the ingenuity that
stood. Basic organic molecules that compose they show in this domain, and even more in
blue pigments naturally absorb in the ultra- the extraordinary saving of means to develop
violet, the violet, and the blue, making them their various tricks. Indeed, if through our
appear as complementary colors: red, orange, scientic approach, we distinguished the main
and yellow. However, there are no clear reasons coloration processes, no buttery actually uses
why structural colors remain conned to vio- one to the exclusion of another! The effects are
let and blue. We are forced to notice that in this almost always combined in order to obtain,
case, exceptions still exist, but they are very few. with only a few basic elements, the innite
Nevertheless, this splitting of the visible spec- palette of colors that one can observe. Why
trum, in addition to the fact that it offers great develop a green pigment when it is possible
saving of means, enablesas we will seethe to obtain the color by mixing an interferential
creation of composed colors, of mixed origin, blue or a blue of scattering with a pigmentary
one of the compounds of which (physical) is yellow, a very common melanin? In the same
iridescent and the other (pigmentary) is not. way, Morphos, the undisputed champions of
This results in remarkable and subtle colored interferential colors, distinguish themselves by
variations. combining pigments that form a background
Even if it is long, difcult, and incomplete, more or less opaque and accentuating or, on
the study of pigments has apparently been an the contrary, diminishing the their vividness,
easier task for the scientistand not for the but- sometimes even annihilating it by curving its
tery. The chemical analysis requirements are wings, in order to obtain the dull blue of a mere
decreasing but still involve a substantial quan- Argus?
tity of material. The total of the scales of a big Butteries are often considered to be the
Pieridae does not weigh more than one to two standard bearers of insects, decried crea-
milligrams, from which the pigment only rep- tures that sometimes provoke brutal reac-
resents a small percentage. So, not surprisingly, tions among us. Butteries, at least diurnal
it required some 215,000 butteries in 1925 to species, are exempted because of their evanes-
extract . . . 39 grams of pigment. cent beauty. The color and the texture of their
We have just briey exposed the different wings, which remind us of the most precious
means used by butteries to adorn themselves fabric, play an important role in our infatu-
with such wonderful colors, and conversely, to ation. It is actually to answer the questions
make themselves invisible to our eyes. So but- of a renowned textile manufacturer that this
teries actually use all the processes of color research project was launched at the Labo-
creation that are available to living beings. ratoire dOptique des Solides of the Univer-
They are barred from some other sources, es- site Pierre and Marie Curie in Paris. How can
sentially colors created by the crystalline eld one manufacture, without using any pigment,
About the Book 5

a colored ber to create fabric and garments cles, wings, and structures of butteries and
with the same iridescence as butteries? We Coleoptera. If, once hatched, the imago keeps
had just entered a fascinating and extraordi- its colors and motifs forever, the latter under-
narily complex world. Yet, the richness of the goes changes at some crucial stages of the insect
explanations was as fascinating and other po- life. These critical stages will be described in
tential applications were found, which were those chapters. I focused on the scales and cuti-
even more astonishing. I will mention these cles of the insects where colors primarily dwell.
biomimetic applications of butteries in the In this part, I presentedquite synthetically
correspondent chapters, which will lead us given their diversity and with a great num-
quite far from the little itting buttery, so dear ber of illustrationsthe arrangements, shapes,
to Lamartine. and ne structures of the scales. The colors of
insects are changing in many ways. They are
iridescent of course, in that they change accord-
About the Book ing to the point of view or lighting: that is the
subject of the present book. But they can also
I tried to write this book so that it could be change under natural or articial constraints.
read at different levels or at various speeds to Even if this is never the case with butteries
enable everyone to get something out of it. This and hardly ever the case for Coleoptera, I found
book is for numerous audiences, including the it important to shortly mention the structures
following: students who often tackle in a disor- that enable this changing in colors. The study
ganized manner the various subjects gathered of material known as X-chromes, which can
here; physicists or engineers striving for origi- change colors under a constraint X, represents a
nal structures and new applications; and biol- very active eld for civil and military research.
ogists to whom, as I am not a biologist, I have I also present several structural principles and
no pretense of teaching anything but to whom some simple experiments to determine the ori-
I would like to remind or to put together some gin of colors in insects.
notions of physics that may be somehow dis-
tant. This book is also for the inquiring mind or
the artist who would like to go a little beyond
fascination. This book falls into three parts that
are not necessarily related. The main theme
coloris extremely rich and complex as it re-
lates to a variety of elds.
When one has seen a buttery eaten by a
bird and another buttery remain perfectly
invisible to the brushing wings of the same
bird, the reason for being of those colors, so
beautiful and loud at the same time, cannot
but appeal to us. That is why I decided to start
this book by trying to answer this question.
Even before describing the orders that concern
us hereLepidoptera and ColeopteraI want
to address this questionbecause the question
goes far beyond the classication of insects. In
this part, I evoke color as part of the defen-
sive gear of insects and generally present the
various types of mimicry principles. The next Figure 1.5. Iris, gods messenger. Spreading her
scarf, she makes appear the rainbows (detail, Louvre
two chapters, structured along the same line, Museum). Inset: Aparatura iris, the changing Great
mention a certain number of elements, which Mars. One of the rare iridescent blue butteries in
we all know more or less about: the life cy- our countries.
6 Chapter 1 Iridescence

Whatever its origin, color results from an in- tion and modeling might sometimes seem too
teraction between light and matter. This inter- detailed. If the results only concern specialists,
action, addressed in Chapter 6, is the part that it will still provide the inquiring mind with an
is the most distant from the view of the little idea of the secrets of nature and the extent of
buttery. In addition to an inevitably super- knowledge available, partly unexplored.
cial presentation of notions of solid state optics Still in this part on electromagnetic proper-
and index of material, I draw a panorama of the ties of insects, a chapter is dedicated to the
effects of polarization of light and of their treat- radiation qualities of wings. Here, we reach
ments. If insects are vividly colored animals, the limits of the study since radiations are
they also behave as polarizers. Polarization is not comprised in the colored spectrum but in
a chromatic character that we are not sensitive the infrared. Color is nevertheless related to
to and can originate interesting discoveries and this problem as it directly affects the absorp-
applications. This chapter is certainly the most tion of solar energy by dark butteries, or its
complex of the book and if the reader avoids reection towards the body by light-colored
it at his rst reading, he will still have a good butteries.
understanding of the book. Such a reader will Lastly, as a counterpart of the rst chapter on
be able to come back to it afterwards if cer- the reason of colors and after this long devel-
tain exposed phenomena stimulates his curios- opment on their physical and chemical origins,
ity enough that the equations dont scare him the book ends with a chapter dealing with col-
anymore. orimetry: how to name a color, measure it, and
The gratication for such an effort can be how humans and butteries perceive it.
found in the third main part, which describes This book will have achieved its purpose if it
the various types of colors produced by the can stimulate the readers desire to understand
insect, phenomenon by phenomenon. Each of the reason of being of those colors and how they
them is illustrated at all different scales by one are produced. The reason why is a fascinating
or several type-species that uses it primarily. subject. As far as the how is concerned, one
For some of the species and for the require- should just imagine that an equation can be
ments of certain specic studies, the descrip- beautiful too.
2
Why Colors

Very few species in the animal world have such which the best trick is total immobility, which
diversied colored motifsas much in the pat- drives us back to the rst imperative.
tern as in the colors themselvesas butteries. In order to comply with the three impera-
These remarkable variations are the mere coun- tives that can be contradictory, butteries have
terpart of the innite diversity of biotypes. Bio- developed an amazing palette of processes,
types in which adult butteries must within a among which color, or more generally optics,
short period of time reproduce, feed, and eat play an essential and sometimes primary part.
without being eaten. This equates to three im- As we will see later on, pigments spring from
peratives: to move, to recognize, and to escape preexisting metabolic compounds. They are of-
predators. ten toxic and stocked we are tempted to say
for lack of anything betterin wings, which
To Move are dry organs where toxic compounds cannot
Imagos feed mostly on nectar, a mix of water cause any harm.
and simple sugars that specic owers pro- Coloration, and thus the pigmentary func-
duce, which butteries must recognize and ap- tion, is a secondary consequence of the proper-
proach. They must y to feed but also to seek ties of those compounds. Yet, it still plays an im-
partners and reproduce. portant part in the vital processes listed above.
To Recognize All of them are closely linked, even when we
Let us imagine a triangle of senses, the three try to study them separately.
angles that represent of the main sensorial stim- Butteries have to y in order to feed and
uli: olfactoryl, visual, and auditory. On this tri- to reproduce. To reproduce they must make
angle, butteries would, contrary to the ma- themselves recognizable, hence make them-
jority of their predators, be situated on the selves visible. These two vital activities make
chemical/ visual base. Butteries are known them easy targets for predators. What role can
to be quite sensitive to odors and pheromones. coloration play in the strategies designed to
But the chemical messages conceal numerous conciliate the two contradictory activities? This
disadvantages: Their dynamic and variability role obviously depends on the ecologic context
are limited and they depend on uncertain exte- of butteries and predators, which must trap
rior conditions such as wind. Optical recogni- the buttery next to one of the three angles.
tion is therefore vital and very well developed.
To Display
To Escape Predators A great number of butteries are not palatable,
Lepidoptera have no system of active defense either because of their repelling taste or because
but their predators are many and fearsome: they represent a danger for predators (toxicity).
They are mostly birds, reptiles (chameleons, Thus, it is in the interest of these species to show
lizards, and the like), batrachians, spiders, and it and deliver a clear message to predators. That
other insects (for example, mantids) against is why warning colors can be loud and intense.

7
8 Chapter 2 Why Colors

Figure 2.1. Relative part of sensorial channels


among butteries, birds, lizards. Butteries remain
mainly close to the visual point of the chemical/
visual baseline and relatively close to the lizard do-
main. Under the strict point of view of safety, their
interest will be to go further and develop chemical
systems for communication and recognition. Many
of them do it, but none can avoid optical signals.
What is lost concerning safety is compensated by a
much more efcient sexual recognition or a better
management of thermal ux. Tactile channels have Figure 2.2. In a given place, a buttery must be able
been neglected here. both to hide for escaping from its predators and to
display its specic characteristics, sexual or not, in
order to: (i) ensure its reproduction; (ii) push back
predators; and (iii) collect the exact amount of solar
They can play a part in sexual recognition too. radiation indispensable to give it the energy for tak-
Other perfectly edible butteries may also try ing off. These requirements are often contradictory
to enjoy the protection provided by such colors. and the insect has to discover an optimal balance be-
tween these three constraints, the relative amount of
This process of real or faked warning through
which can change. Other complementary strategies
colors represents a defense strategy, mimicry, using nonvisual techniques can be used.
that is developed by butteries in particular.

To Hide are limited and the generated efforts, especially


Butteries perfected the art of visual camou- to get off, are substantial.
age, which is, by the way, widespread in the In order to get off, the average tempera-
animal world. Visual camouage is particu- ture at their thorax, where the extensors and
larly used by animals at rest. Only a specic abductors of the wings are situated, must be
part of the animal is affected: one face, gener- over 36 C. The energy required to reach such
ally the ventral side, or one pair of wings (the a temperature is often superior to the ecologic
anterior ones that overlap the posterior ones at context. It can be produced partly through an
rest). In this way, visual camouage is compat- endogeneous way by muscles and through an
ible with the effects of mimicry as mentioned exogeneous one by picking up solar energy di-
earlier. rectly onto the body, onto the wings if they ab-
sorb all or part of the solar spectrum, or onto
To Pick Up the body after reection on wings (for white
Diurnal butteries are heterotherm insects, wings). Besides, butteries are incapable of y-
which means that they get a relatively impor- ing, or even surviving, if their interior temper-
tant part of their energy from the environment. ature goes over 40 C. The surplus of energy
But, they are quite poor navigators. The results generated during the ight must be eliminated
Variations in Colors 9

through convection or radiation. The energetic The phenomenon is beautifully illustrated


balance essentially depends on the coloration by the case of a Phaleanae: Biston betularia.
of the wings and their radiative properties. Most of adults as well as heterocers are active
The markings of a buttery proceed from a only at dusk and settle on various supports
subtle balance between those three main at- branches, trunks, and so onwith their wings
tractors: perfect camouage (to escape from wide open. In the early 19th century, while the
predators), maximum coloration (for recogni- species was only known to be light-colored,
tion or warning), and the absence of color, some totally dark specimens started to appear
black (to pick up as much solar energy as around Manchester. The phenomenon spread
possible). rapidly and in 1950, the dark species was in
the majority in Eastern England as well as in
Variations in Colors the Glasgow and Newcastle areas; that is, the
highly industrialized regions. That is why the
Variations in colors within a species are phenomenon was called industrial melanism.
widespread. They often proceed from specic We now know that this evolution was not
ecologic conditions, but may also be acciden- caused by the ingestion of soot by caterpillars,
tal. In most of these latter cases, it is related as people thought at the time. It is indeed a con-
to melanism phenomena that can result in the sequence of natural selection, which favored
insect turning totally black. the form that was best adapted to the envi-
ronment. Before the industrial revolution, trees
were covered with lichens on which the beige
veins and white color of the light form of Bis-
ton betularia were perfectly invisible, whereas
dark mutants were rapidly eliminated. When
trunks started to turn black as a consequence
of the soot released by manufactures and the
disappearance of lichens, modalities reversed
and the light form, which had thus become vul-
nerable, disappeared completely.

Figure 2.3. Industrial melanism in Biston betularia.


Melanic forms were rare in the past because they
were more often exposed to predators by their opti-
cal contrast on the light bark of birch trees. Nowa-
days, they are preponderant, due to the darken-
ing of birch trees under the effects of pollution. Re-
cent efforts against pollution make the melanic light Figure 2.4. Example of procryptic colors. Biston
forms reappear, showing thus the dynamism and strataria (5 cm). Here, as often, shades and chromatic
efciency of adaptation. motifs of the background are imitated.
10 Chapter 2 Why Colors

This widely accepted interpretation corre- isokinemy, that is the same mobility as their
sponds to the decline of the dark form noticed models. The objects that are the more often
over the last decades after anti-pollution mea- imitated are leaves and bark but also include
sures were launched. lichens and bird excrements. We wont address
the genesis of mimicry and camouage; sev-
eral theories were offered that involve natu-
Colors: Another Classication ral selection in the origin, dynamic, and sta-
bility of the phenomenon. These questions are
As regards selection, we can try to classify col- still unanswered today and we will only briey
ors not according to their physical character- set out the modalities of the main types of
istics or physico-chemical origins, but to their mimicry.
role in the strategy for survival and develop-
ment of a species. We can, as the English nat-
uralist Alfred Russel Wallace did in the late Mimicry
19th century, distinguish colors between those
designed to warn and those for camouage. As is often the case, the fathers of important
The latter allow the insect to become invisible discoveries tend to remain unknown from the
in its environment for defensive or offensive public, whereas their work comes at the right
purposes. As we have seen, there is a price time to support great theories with a more uni-
to pay, which is a decrease in mobility. On versal impact. Henry Walter Bates, Fritz Muller,
the contrary, warning colors, which are very and on another hand, Charles Darwin, illus-
loud a priori, are related to a real danger or a trate this.
faked one from the viewpoint of predators. A H.W. Bates, an English naturalist and ex-
few years later, Sir Edward B. Pulton offered plorer, was born in Leicester in 1825 and
another classication by distinguishing colors died in London in 1892. He explored the
that deceive from those that dont. The latter Amazon valley for 11 years and came back with
warn against a real danger: toxicity, venom,
spine, repelling taste . . . These colorations are
called sematic colorations.
Deceiving colors, called apetetic, include all
hiding or cryptic colors and warning col-
ors faking a danger or pseudo-sematic. Each
branch consists of other ramication depend-
ing on whether the insect carrying colors is a
prey (procryptic) or a predator (anti-cryptic),
whether it tries to repel (pseudoaposematic),
or, on the contrary, to attract (pseudepise-
matic). In this latter group, which is actually
quite rare within the buttery population, the
widespread cases of auto-mimicry, in which
the motifs of a nonvital body part (like poste-
rior wings) fake those of the vital part (the head
or anterior wings): Those are called parase-
matic colors.
While sematic colored butteries, which
have nothing to hide, have a wide range
of colors and patterns, apathetic must re-
sort to isochromy with their model or with
a widespread element of their environment Figure 2.5. W. Bates, lithography by W. Purkiss
(homotypy) and to isomorphy, and even (1880). C. Hulton Picture Library.
Mimicry 11

with the Church, he was kidnapped, and his


family diedyet his life proved exceptionally
rich.

Basic Principles of Mimicry


Mimicry is an intricate biologic phenomenon
that aims at ensuring protection of the species
against predators (repelling mimicry), which
concerns Lepidoptera. Or it can also aim at pro-
viding the predator with an advantage over
his preys (attracting mimicry). We will focus
on the optical strategies involving colors in the
rst type only. Chemical or auditory actions can
obtain similar effects. Mimicry involves at least
three protagonists, among which the predator
is either duped by the deceiving appearance
of an eatable prey (Batesian mimicry) and the
predator is thus described as an effective dupe,

Figure 2.6. F. Muller in Brazil in 1890.

14,712 species of insects, among which 8,000


were unknown and for this, the king awarded
him 800! Even more than Bates, Fritz Muller
is the scientist of the 19th century in all its
glory. Born in Windischlolzlausen (Thuringe)
in 1822, he had an extraordinary gift for lan-
guages and spoke Italian, Russian, Syrian, Ara-
bic, English, and later, of course, Portuguese.
He studied pharmacy, mathematics, zoology,
and medicine at the same time. He was forced
to ee his country because of commitments
during the 1848 Revolution. His anticlerical
and libertarian ideas were unacceptable in the
then-intolerant Prussia. He exiled himself to
Brazil where he settled down as a farmer and
then as a mathematics teacher. Attached to the
National Museum of Rio with the ofcial title
traveling naturalist, his great talent and his
genius as a naturalist, his modern and precur-
sory mind give him a place among the greatest.
Darwin, who knew his subject and whom
Mullers observations provided with weighty
arguments, predicted honor and recognition Figure 2.7. Some classical examples of batesian
mimicry . . . Danaus chrysippus (6 cm) non-edible (top
for his work . . . posthumously! Muller died in left) and one of its numerous mimickers, Hypolimnas
Blumenau, Brazil, in 1897 after a tormented missipus (top right), Tirumala limniace (9 cm) (middle)
and difcult lifehe was in a constant battle and its mimicker, Chilasa clytia (10 cm) (bottom).
12 Chapter 2 Why Colors

or it is warned against a real danger by means of would annihilate the warning effect of sematic
a minimalist signal, like a color, common to dif- colorations of the latter. In this way, Batesian
ferent uneatable species (Mullerian mimicry). mimicry is similar to some parasitism from
Here there is no deceiving in the proper sense, which a model can escape by changing its
so the predator is described as an indifferent markings. This can lead to a parallel evolution
dupe. The two strategies require a geographi- of the mimic and the model, the latter becom-
cal concordance of the various protagonists. ing more and more different from the former,
which, put in danger, always tries to resem-
ble it.
Batesian Mimicry
This is certainly one of the most remarkable
Mullerian Mimicry
protective strategies of Lepidoptera. Here, the
mimic, an eatable buttery, imitates the mark- Species that are truly protected because they
ings and ying manners of an uneatable in- are uneatable or dangerous and present se-
sect for whatever reason: toxicity, repelling matic colors are not ipso facto safe from preda-
taste, or venom. This kind of mimicry can be tors. The danger here occurs during the learn-
both intra or inter order/group. Swamps ing stage of the predator, which associates an
Hymenopteraare often imitated by Lepi- annoyance to a warning signal only after exper-
doptera. If it guarantees a relative immunity imenting it many times. This is often deadly for
to the mimic, Batesian mimicry puts the model a buttery, yet the risk gets smaller according
in danger and its efciency depends on the ra- to the clarity and non-ambiguity of the mes-
tio between the mimic population and the imi- sage. The latter must be strong. As sematic
tated one. Predators learn their lessons through colorations are loud, they should not be too
successive successes (eatable insects) and fail- many. Thus, protected species should always
ures (uneatable insects); if the proportion of present the same markings to predators. In-
the former outnumbered that of the latter, it deed, they reduce and share the chances of

Figure 2.8. Mullerian mimicry in Heliconius genus.


Mimicry 13

Figure 2.10. Guillauminian association, or ash


color: Catacolia ilia (photo A.B. Klots).

colorations. Here, the spotted insect suddenly


reveals the vivid colors of its posterior wings
by getting off, deceiving the predator for a
moment.
Throughout our summary description of
Figure 2.9. Examples of Mullerian mimicry. Itu- those theories, we considered the pressure of
lia hilione (9 cm) (top), Thyridia confusa (9 cm) predators as a sole and unique constraint. The
(bottom).
logical consequence of this would be a great
standardization of colors in a given geograph-
being eaten during the learning stage and ac- ical area, through Batesian, Mullerian, or any
celerate it. This kind of association, known as other mimicry, or even due to hazard. But the
Mullerian, pushes the limits of mimicry be- experience refutes this and one can observe
cause it implies that everybody imitates every- in the same place butteries from the same
body and so there are no denite models or species, threatened by the same predators and
mimics in the end. Furthermore, if any eatable all uneatable, presenting different markings.
intruder enters the association, as a result of This also contradicts the Mullerian theory! One
hazard or Batesian mimicry, it reduces the im- can also encounter cases of Batesian mimicry
pact of the message. concerning one gender only or a species pre-
To end this chapter about the protective senting polymorph mimicry, which is imitat-
role of colors, let us mention Guillauminian ing various models. This is due to the other
associations[correct word?]les associations constraints that are as vital to the species or
dintrouvables, to take up Michel Boulards the insect as recognition of genders and ther-
beautiful expressionwhich include sematic mal regulation. If one of them is slackened
and procryptic colorations. The predator, at- because the species lives at the top or at the
tracted by the strong visual signal of its base of the canopy and therefore receives a dif-
prey when it is ying, but frustrated when ferent solar radiance, for instancethe whole
it becomes invisible at rest, eventually as- balance is affected, either upset or displaced
sociates sematic colors not to an uneatable to another point, which is less of a constraint
prey anymore, but to the absence of prey! for the two others. New markings will then
The reverse strategy also exists with blitz emerge.
3
Lepidoptera Description and
Scales of Observation

Our present purpose is not to write a book on We can dene other big groups that are re-
butteries, but rather a book on their colors. It is lated to the previous ones and usually com-
still useful to begin by situating butteries into monly known as micro and macro Lepidoptera.
classications and review some general facts The latter group is composed of the most devel-
concerning their anatomy, development, and oped diurnal and nocturnal butteries, while
the structure of their wings. former group includes the less developed and
Among insects, a group that already out- usually small nocturnal butteries.
weighs the animal world by its remarkable
diversity, ve orders outclass all of the oth-
ers by the number of species composing Outlining a Classication
them. Those are Coleoptera (300,000 species),
Hymenoptera (250,000 species), Diptera and Distinctions are sometimes more marked in
Hemiptera (150,000 to 200,000 species), and English. Indeed, the word butteries comprises
Lepidoptera with about 170,000 species. It is all of Rhopalocers, the real diurnal butter-
mere utopia to try to establish a comprehen- ies, and moths, the Heterocera, mostly noc-
sive classication of such a huge number of turnal butteries with feathery or threadlike
species, all the more since only 5 to 20 percent antennas. These terms designate some use-
depending on ordersof the total have of- ful categories, the limits of which still remain
cially been described and studied. blurry. However, they are quite acceptable
The order of Lepidoptera comprises diur- for us.
nal and nocturnal butteries. The word derives Heteronera are subdivided into a large num-
from the Greek term Lepidos (scales) and Pteris ber of superfamilies, which will not be de-
(wings), and therefore means scaled wings or scribed here. Indeed, the majority of butter-
wings covered with scales. This is one of their ies presented in our study come from one
main characteristics, and yet some of them only superfamily only, the Papilionoidea, consisting
partially present scales while others are totally of 14 families, with one well-known excep-
deprived of them. tion, the Uranoidea, illustrated by the famous
The large majority of nocturnal butteries Uraniidae. It is impossible to describe in a few
and the total of the diurnal belong to the in- lines insects as diverse as butteries, which are
fraorder of Heteroneura, which means that adapted to almost all environments, and the
they present different veins on their anterior size of which varies from 2 mm for the mi-
and posterior wings (Figure 3.9). Homoneura nuscule Nepticulides to more than 20 cm for
consists of some primitive butteries present- the big saturnides. We will only briey mention
ing the same veins on both wings. We will their development and characteristics. Butter-
hardly mention them, as they have little or no ies appeared in the early Jurassic, 200 million
coloring. years ago, and despite their apparent fragility,

14
Summary Description of Butteries 15

they survived all the cataclysms that proved


fatal to many a species.

Summary Description of Butteries


The development of butteries from egg to
adulthood consists of four very different
stages:
r The egg or embryonal stage.
r The caterpillar or larval stage, which is essen-
tially a development stage.
r The chrysalis or pupa, which is a phase of
reconstruction and differentiation of tissues.
r And at last, the imago (adult) buttery,
which is sexually mature and capable of
ying.
We will describe those stages when they
can inuence the motifs and colorations of the
mature insect, which concerns the larval and
nymphal stages. Figure 3.2. The life cycle of Lepidopterons. Here Pa-
pilio machaon. Certain stages of the buttery life are
Caterpillar particularly critical for its color.

Like the adult insect, the caterpillar consists of


the head (which has ve combined segments),

Cremaster

Wings

Antennae

Eyes
Figure 3.1. Schematic representation of a caterpillar
highlighting the main anatomic partsArcherontia
atroposthe deaths-head moth. A caterpillar can Figure 3.3. Schematic representation of a Melitea
look quite different from one shedding to the phoeba chrysalisJapan. The Chrysalis is the rst
other. More exposed than the imago because they specialized stage of adulthood including all the fu-
move relatively little, caterpillars possess a sub- ture organs. Compound eyes are perceivable, while
stantial defense arsenal, mainly camouage and antennae, proboscis, wings and legs are along the
mimicry/mimetism like the imago, but also chemi- body, distinguishable in relief. The external cuticle
cal means. can be highly colored.
16 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.4. Schematic representation of a


Rhopalocer. One can note on the head, un-
der compound eyes, the folded labial palpes
and proboscis. Anterior legs are atrophied and
thus unable to move among many Rhopalo-
cers. Each of the rst eight abdomen segments
is equipped with a pair of respiratory stigmas.
At the posterior end, the anus and reproduc-
tive organs (genitalia).

(a)

Figure 3.5. The structured membrane of


Grapium weiskeipurple areastop view on
the left, and cross on the right. This structure
is similar to that of green areas where a biliary
pigment was found. The pigment is diffuse in
both cleaved membranes, which are well vis-
(b) ible on the picture.
Summary Description of Butteries 17

Figure 3.6. Morpho menelaus


alary membrane. Most of the
scales were removed on both
sides. One can distinguish the
two membranes of the wing,
and on the inferior membrane,
the scales pedicels joined by a
thin tracheole.

Figure 3.7. The different


scales used to observe but-
tery wings and their units
of measure. The entire wing
(5 cm), scales (100 m), striae
(10-1 m) and striae ne struc-
tures (1 m100 nm).
18 Chapter 3 Lepidoptera Description and Scales of Observation

in the xation of the chrysalis. The head is pro-


vided with a pair of small antennas, two single
eyes situated on both sides of the head, and
large mandibles. The larval stage is essentially
a development stage consisting of successive
sheddingbetween 3 to 7 times depending on
species. Caterpillars primarily feed on plants
and their digestive systems are capable of elim-
inating natural toxic substances or stocking
them. Some of them can make pigments out
of toxins, which are yet absent in adult insects,
with the possible exception of some yellow or
Figure 3.8. Optical microscope cross-section of an cream-colored vegetal avones.
Ornitoptera priamus poseidon wing. The alary mem-
brane is in the center, surrounded by the various
layers of scales. The exterior scales here are the struc- Chrysalis
tural, while those closer to the center are pigmentary.
The pupal stage, or chrysalis, is a stage of the
adult state in which all of the future organs
the thorax (which is composed of three seg- of the imago are already in place. Some of
ments, T1 to T3), and the abdomen (which is them can be perceived through the integument.
composed of ten segments, A1 to A10, that are We wont describe the nymphal diapause since
not always distinguishable). it doesnt really concern us. However, a cer-
Each of the thoracic segments presents a pair tain point directly concerns us. One can eas-
of articulated legs composed of ve segments ily understand that it is during this stage of
and ended by a claw and which are distinct complete reorganization when the insect actu-
from pseudopods or membranous legs, which ally takes on its nal form, that the patterns and
are not articulated and are situated on seg- colors of the wings of the imago are dened
ments A3 to A6 on the abdomen. The latter and that exterior constraints can change those
end is a contractile sucker crowned with hooks. plans. There are actually very few exchanges
Pseudopods play a very important role in the between the chrysalis, protected as it is by a
mobility of the larva and are sometimes used quasi-waterproof integument, and the exterior.

Figure 3.9. Heteroneura stan-


dard venation and nomencla-
ture of the edges, angles, and
of the main wing areas (a). In
comparison, a schematic rep-
resentation of an Homoneura
venation (b).
Summary Description of Butteries 19

nize and encounter sexual partners. This is also


the stage in which, through his search for food
and partners, the insect is the most exposed to
dangers.
Like the larva, the body of an adult butter-
y comprises three distinct parts: head, thorax,
and abdomen. Each part is covered with scales
of various shapes and functions, most of which
dont concern us. We will focus on the thorax
and, to a lesser extent, the head and its two ex-
traordinary composed eyes.
As in every insect (and their larva), the tho-
rax is composed of three united segments to
which the three pairs of legs are attached: pro-
Figure 3.10. The basic Nymphaloid plane and the thorax, mesothorax, and metathorax. Anterior
different symmetry systemsaccording to Siiffert, wings are already present in an embryonic
1925.
state in the caterpillar and in an almost nal
state in the chrysalis. Anterior and posterior
The chrysalis does not feed nor does it excrete wings are linked with a passive device: the
anything. Respiratory exchanges and evapora- frenulum. The thorax also contains the mus-
tion are minimal. Only thermal variations and cular groups that are responsible for ying and
hygrometric, natural or experimental ones to walking.
a lesser extentapplied to denite stages of
diapausehave an impact on the motifs and
colors of the wings.
Moderate and prolonged thermal variations
tend to bring out motifs, colors, and seasonal
forms that are usually distinct within a same
species (seasonal dimorphism). More spectac-
ular modications can be obtained experimen-
tally by subjecting the chrysalis during the crit-
ical phase of the early nymphal stage to high
variations in temperature. Even more interest-
ing is that in 1933, a detailed stratication of the
critical phase was established: an intervention
altering the motifs during the rst sub-phase,
the structures and development of scales
during the second sub-phase, and nally their
pigmentation during the third one. Thus, ac-
cording to species, the nal aspect of the motifs
and colors of the imago are dened during
the second or third day after the chrysalis was
formed.

Imago
The imago, the last stage in the develop-
ment of the buttery, essentially concerns
reproduction. The imago is mobile and pro- Figure 3.11. Series of dislocations on the ventral side
vided with all the organs necessary to recog- of a Charaxes jasius anterior wing.
20 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.12. Scanning electronic micro-


scope image of Parnassius apollo ocellus.
And its image by using a binocular lens
glass at approximately the same magni-
cation. One can note the scales color
variations as well as their correspond-
ing structures.

Description of the Wings hundred of micrometers, m), then that of


striae (roughly micrometers), and lastly, that
We are going to describe the wing of the butter- of the structure of striae or inter-striae spacing
y and its optical properties through two paral- (from around 50 to 100 nanometers). Each scale
lel yet opposite points of view. As we will see, corresponds to a specic means of observation
we can distinguish ve levels of observation and measurement, and also to its role in optical
or ve magnications, ranging from macro- properties! It is logical to describe the anatomy
scopic scale, the whole wing measured in cm, of the wing by zooming in, from the macro-
to the molecular scale measured in nanome- scopic to the microscopic, each element charac-
ters (nm). Between the two and in descend- terizing one scale supporting its smaller com-
ing order are the scale of scales(roughly the ponents. As for optical properties, if the scale is
Description of the Wings 21

Figure 3.13. Morpho godarti ocelli, ventral side. On the left image, the ocellus is spherical and its concentric
circles present the same thickness. On the right image, the ocellus, in the neighboring cell, starts elongating,
the circle thickness increasing towards the wing apexlocated on the right of the images. Blue scales,
which are almost absent in the spherical ocellus, cover half of the central motif in the elongating ocellus.
Generally conned in an alary cell, ocelli can sometimes substantially expand onto the neighboring cells
(Caligo idomeneus).
22 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.14. Circular and ellipsoidal ocelli


formation process resulting from the com-
bined action of two components. The in-
hibitor diffuses on the surface of the overall
wing from the base to the peripheryon
the frontwhereas the activator diffuses
from the foci. At the rst focus, the in-
hibiting gradient is equal to zero, result-
ing in a circular ocellus. It is decreasing to
a great extent at the second focus, leading
to an ellipsoidal ocellus. Finally, the activa-
tor never reaches the critical concentration;
therefore, the ocellus is not formed.

the same, we will study it in the opposite direc- remarkable surface structure that scarcely af-
tion, in ascending order. Zooming out, color is fects coloration, but that slightly diffuses light
created at the bottom of the scale and is modi- and gives a pale, velvety aspect to these areas.
ed at each rung to eventually obtain the nal One can roughly imagine the wings of an
macroscopic effect. ideal Lepidoptera as two triangles, the an-
gles of which are more or less rounded off. The
angles, sides, and main concentric areas have
Macroscopic Level: Structure of the Wings been designated with different names, which
Lepidoptera possess four membranous
wings covered with relatively few veins in
comparison to that of other insects. The wing
venation characterizes species and thus allows
one to classify species.
The wing consists of one double membrane,
the two of which developed independently and
united at the nal formation stage of the wing.
That is why the colors, motifs, and implanta-
tion of the scales of the ventral and dorsal sides
of the wing are totally independent and often
highly contrasted. Already developed in the
chrysalis, the double membrane is covered by
tracheas. During their development, the two
membranes join each other and at the same
time, are sclerotized along tracheas forming the
veins, which continue to transport the nervous,
gas, and haemolymph messages.
These membranes are the support of the
scales, the surface on which colors mainly
reside in Lepidoptera. Diffuse pigments can
nevertheless be found on membranes, like
that found with Graphium (see Chapter 11). In Figure 3.15. Schematic representation of a Lepi-
this specic case, the membrane consists of a dopteron scale.
Description of the Wings 23

Figure 3.16. The base of a Dixocopa chlotilda scale (a)


showing the insertion of scales in the peduncle.

is quite convenient for the study of the wing


(Figure 3.9.). We will only represent the princi-
pal divisions here, which is more than enough
to illustrate our study.
The ratio between the size of the wings and
that of the body is a critical parameter con-
cerning the study of the ying process but also
when it comes to thermal exchanges, which
particularly concern us. This ratio is greatly re-
duced among species that present a high ying
speed, like Sphingidae and substantial among
Nymphalidae, which often glide.
The variations in shape and size of the wings
between species and also within a given species
(b)
between male and female, are innumerable, so
that some of them can not rank in our diagram. Figure 3.17. Transmission photon microscope view
They can present tails, sometimes very long, of an isolated Morpho menelaus structural scale, top
they can be oblong and without denite anal or view. The scale presents a very geometric shape but
internal angles, or they can even form narrow is not perfectly plane. One can note the same irides-
fringed lobes. Let us now mention the dispar- cent effects, from blue to purple, as on the whole
wing. Below, Archeoprepona pigmentary scale; striae
ities that sometimes exist between males and are more spaced out, and the scale apex is character-
females within one species. Whereas among istically multi-lobed.
numerous species, males are smaller than fe-
males, there are many contradictory cases of
partial or total aptery affecting the latter only. various shapes of wings, which would go be-
Aptery can be found in specic biotypes (in yond our subject. Some will be included in our
an island environment) and protects the insect illustrations, thus outlining their extraordinary
from being carried away towards the sea or diversity.
away from the eggs by winds. All the transi- As we have already said, the pattern of veins
tions between well developed wings and com- characterizes species. Several designation and
pletely atrophied ones exist, and even when numeration systems are available. We chose
there is the total absence of a pair of wings. We one only to situate more precisely measures.
wont give a comprehensive description of the We chose, without any a priori on its relevance,
24 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.18. SEM view of Par-


nassius apollon scales. Located in
a white area of the wing, these
scales are little inclined on the
membrane and strongly curved,
thus revealing its two sides. The
superior membrane is closely and
regularly striated, whereas the in-
ferior one, turning outward here,
presents larger and less distinct
structures.

Tillards system, which is more adapted to des- joining the radial and the cubital, and from
ignate triangles than the Anglo-Saxon numer- which proceed the medial ones: It is the dis-
ation. The majority of secondary veins actually cal crossvein. But of course, scarce are the but-
proceed from two main big veins, known as teries that match this diagram, even without
radial and cubital. Their subdivisions are num- counting the supernumerary, merged, or sub-
bered from the rear to the back. The previous divided veins . . .
ones are the subcostal, and those between the The areas delimited by the pattern of veins
two medial and the following ones are the ab- and the edge of the wings, known as alary cells,
dominal. The main veins are all oriented from are more important to us. They were named af-
the base towards the periphery of the wing, ex- ter their inferior vein, except for the triangular
cepted for one transversal vein, the only one area delimited by the radial, cubital, and discal

Figure 3.19. SEM image of the


end of a Pieridae androconia
scale, Pieris rapae. These special-
ized scales give out pheromones.
They are here spread all over
the wing. Among many species,
they form distinct bunches in a
few specic places, like the ab-
domen extremity, or they can
also be arranged in rows alternat-
ing with cover scales. Androconia
scale cuticle presents numerous
pores responsible for evaporation
secreta.
Description of the Wings 25

vein: the discal cell. Optical measures focus for


the most part on the biggest areas, Cu1, Cu2,
1A, and the discal cell, as they are relatively at
and homogeneous.

Coloration and Motifs


Colored motifs on the wings are various: eye-
spots, lines, simple or composed strips, ocelli
that are spherical or not. They can be found on
one side only, or on both. A perfect mess or
a great uniformity, this apparent anarchy con-
ceals a pattern, a general outline that is always
more or less transcended from several simple
Figure 3.20. Androconia scale of Hebonia glocippe geometrical transformations.
(Pieridae), surrounded by two typical scalesthe General outlines or basic sketches were de-
superior row scales have been removed in or- veloped by Schannitsal in 1924 and Siifert in
der to expose the peduncles of the two scales 1925 concerning certain groups, such as the
types.
Nymphalidae, one of the very large families
of butteries.
Here are the main characteristics of the
sketch:
r The edges of the wing present a row of ocelli,
a group of concentric circles, centered on the
median axis of each cell.
r Symmetries in relation to median lines of an-
terior and posterior wings. The median line
crosses several groups of colored strips in
their middle, the main of which is situated
in the center of the wing. The latter is the cen-
tral system basing symmetry and includes in
the middle a vividly colored strip: the discal
spot. One can also nd two eccentric systems
of symmetry on the base and the periphery
of the wing.

No buttery species perfectly correspond to


this sketch, yet all of them develop their par-
ticularities based on it. The most common and
obvious transformations are pure and simple
eliminations of all or part of its elements but
also the contraction or dilatation of strips or
ocelli and nally the dislocation, which means
the displacement of a motif along the veins
within a given area.

Formation of Chromatic Motifs


Figure 3.21. Shapes and arrangements char-
acterizing structural and pigmentary scales Chromatic motifs of butteries wings tend to
Archaeoprepona. be very intricate, spreading on several cells,
26 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.22. Transmission opti-


cal microscope image of Morpho
menelaus dorsal side structural
scales immersed in an index
liquid. The ventral side scales
have been removed, but cover
scales, deprived of pigment, are
visible. One can thus only per-
ceive the brown pigments on
the inferior sides of the struc-
tural scales.

even sometimes on the whole wing, and it along the median axis of alary cells, depends
proves quite difcult to experimentally deter- on a specic property of the cells of the motif
mine its biogenetic origin. This is not the case center. Known as the focus, it has been possi-
where ocelli are concerned, which consist of rel- ble to experimentally shift this group of cells
atively small motifs that are always precisely leading to the formation of an ocellus on the
situated. place of the shift, or to remove it, which brought
An ocellus is characterized by a specic pig- about the complete disappearance of the ocel-
mentation and an arrangement of the scales lus. These experiments also establish a rather
contrasting with its environment. The ocellus detailed chronology of the development of the
itself generally includes several colors, yet it ocellus and show that the motif is denite about
does not affect the morphology of its scales. It three days after the wing has been totally pig-
is an established fact now that the arrangement mented, which occurs within the 24 hours be-
of ocelli, which are always placed precisely fore hatching.
Description of the Wings 27

Figure 3.23. The two arrange-


ment types characterizing struc-
tural scales. Above, Morpho-type;
structural scales form the un-
der layer, the ground scales, so
that cover scales are translu-
cent. Below, the reverse situ-
ation in Archaeoprepona. Cover
scales generate structural col-
ors, whereas the ground scales
pigmentaryform an opaque
screen that absorbs the light that
has not been reected by the
cover scales.

The foci are thus a source of information substance, an inhibitory, which allows the syn-
indicating the place where the ocelli can de- thesis of the melanin only when the concen-
velop. The geometry of the motifs of the ocel- tration in the activator is superior to that of
lus can be explained by a double process of the melanin. The fact that the nonspherical
activator-inhibitory kind; this is the theory of or composed ocelli remain symmetrical in re-
the two gradients. lation to the median line of the alary cell
The focus diffuses a substance, an activator implies that the foci (or focus) of this sec-
the rst gradientleading to the synthesis ond substance is situated at the base of the
of a dark melanin after its concentration has wings and that the inhibitory diffuses mostly
reached a critical limit. Under this limit, the towards the peripherythis is the second
scales synthesize another pigment, a yellow gradient.
papiliochromis for instance. This limit may The respective shapes of the two gradients
be determined by the presence of a second allow us to reproduce most of the motifs thanks
28 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.24. SEM image of the ar-


rangement characterizing scales that
produce color by scattering light.
The disorganization doesnt affect
the resulting color, which is rather
lusterless. Overlapping rate is high.
Polyommatus icarus.

to few parameters. An isotropic scattering from here could result from the fusion of peripheral
the foci would thus create circular motifs if the rings of large ocelli.
second gradient were absent (which entails a Gradients of activator and inhibitor concen-
uniform concentration on the whole surface of tration seem to be the key to the system of ar-
the ocellus) and elliptic motifs if the gradient is rangement of the colored motifs of buttery
strong with the main axis towards it. wings. However, the reality is obviously more
It is quite possible, but not yet established, complex and may imply more than two com-
that the motifs of the symmetric systems are ponents (inhibitor of inhibitor) and saturation
also related to rows of foci running across the phenomena of one component with the emer-
wings from the edges to the interior and ab- gence of a limit.
dominal edges. As a plane wave can be split up Finally, the detailed chronology of chemi-
into overlapped spherical waves, straight lines cal reactions turning tyrosine into melanin or

Figure 3.25. The colors produced


by interference and diffraction re-
quire a very regular arrangement of
the scales. SEM. Ornitoptera priamus
poseidon.
Description of the Wings 29

Figure 3.26. Iridescence effects can be substantially weakened by the scales geometry and arrangement. The
most remarkable effect results from at scales that are little or not at all covered by cover scales, like in Morpho
menelaus or Morpho cypris. In Morpho anaxibia, on the right, structural scales are slightly convexhardly
affecting interferential effectsbut are completely covered by highly scattering cover scales. Iridescence
remains, yet the buttery looks dull, compared to the bright Urania Chrysiridia madagascariensis, the structural
scales of which are highly curved but not coveredon the left.

papiliochromis represents a temporal indica- tennas, where they play sensorial, respiratory,
tor that could explain many details of the poly- and odoriferous roles. To make our description
chromatic structure in a given motif. more simple and before describing scales in de-
tail, we would like to distinguish pigmentary
Scales: Structures and Morphologies scales and the colors they produce from the
structural ones, the color of which primarily
Nomenclature Attempt
originates from physical effects related to the
If we focus on the scales covering wings, one ne structure of the membranes. The caterpillar
should remember that the phaneres recover the already shows embryonic wings as tiny spots.
whole of the body, including legs, head, and an- The imaginal discs split up into vesicles during

Figure 3.27. Hebonia glaucippe vossi


Pieridae pigmentary scales. Pig-
mentary scales tend to be lanceo-
late or multi-lobed. Striae are big
and regular.
30 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.28. Cithaeria menander Pili-


form scales. These very long scales
that can be 1 mm long are often con-
centrated in certain places on the
wingbasal and abdominal mar-
gins. They can also be scattered over
the whole wing.

the successive shedding. At last, during the On the surface of the scales, an epicuticular
nymphal shedding, the latter evolve to form the skeleton will emerge that will produce the vari-
pterotecae. Onto the folds of these future wings ous structural systems generating physical col-
wrinkled up in the chrysalis, some cells differ- ors. In the same time, pigments will be stocked
entiate turning into trichogenous cells on one into the epicuticale system or outside of it as
hand and tormogenous on the other hand. The ovoid granules.
former will produce attened and altered hair, The general morphology of the scales is as
scales, and the latter will create scales peduncle. fascinatingly diverse as their roles are. We

Figure 3.29. Piliform scales of


Ornitoptera priamus Poseidon.
Description of the Wings 31

Figure 3.30. Below: Over the


whole anterior wings and
most of the posterior, Cithaeria
menander scales6-cm longare
completely atrophied revealing
the alary membrane, which is
slightly pigmented and produces
thin lm interferences in places.
The only characterizing element
that the insect still presents is
two big ocelles at the posterior
wing ends and a wide scattering
brown area. In the background,
Kramers-Kronig relations that
link the imaginary and the real
part of the dielectric function
of a material and are used to
determine its optical index.

propose to give an outline of the former striae that are visible through an optical mi-
which has never been completedin order to croscope, and on other scales, a secondary pat-
establish the nomenclature. tern of transversal striae. Other, even smaller
The scales of Lepidoptera usually consist of structures can be perceived that are shut in
two chitineou layers that form the limb. Facing those patterns. The lamellae composing the
the alary membrane, the inferior side is gen- two sides of a scale are joined inside by vertical
erally smooth or slightly undulating, whereas trabeculae.
the superior side, which is turned outward, is As we will see, scales are implanted in a
thicker and shows a great diversity of struc- more or less regular manner into the cuticle
tures. Almost all of the scales present a regular of the alary membrane, which is curved as a
pattern of more or less spaced out longitudinal cupule with angular edges, the axis of which
32 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.31. Thick and irregu-


lar striae of a Morpho cypris
androconia.

inclines towards the apex of the wing. Scales Peripheral scales are usually quite long,
present a great variety of shapes as shown sometimes pilliform, but they can also be
by the illustrations in this book: at, concave, haphazardly scattered throughout the wing
undulating, threadlike, pilliform . . . These var- among the standard scales. The end of the limb
ious shapes inuence colors and must be con- can present extremely various shapes: curved,
sidered when analyzing the colorimetry of dense, or even lamentous.
the wing. Within a species, the shapes of
scales greatly depend on their central or pe- Pigmentary or Structural
ripheral implantation on the wing, and also Scales are the center of colors among butter-
sometimes on the implantation of one single ies. We will rst make a distinction concern-
cell. ing the origin of these colors.

Figure 3.32. Grating of struc-


tural striae, above, and of pig-
mentary striae belowat the
same magnication, by using
SEM, Morpho menelaus. Struc-
tural striae, very dense, but also
much higher than the pigmen-
tary, can tilt and interweave in
places.
Description of the Wings 33

Figure 3.33. SEM image of a Morpho menelaus ventral side pigmentary scale. The key to Lepidopterons
colors is concealed in this extraordinary and mysterious architecture. The various elements composing the
thin structure of scales can all be seen here; keystone supporting striae, lamellae partially overlapping one
another along striae and held by a very regular network of microtrabeculae. All scales, whether pigmentary
or/and structural, use these elements to differentiate. Here, the pigment can be found in a diffuse state in
the alary membrane scales, but it can also be found in others as small granules hanging from keystones.
The lamellae, little developed here, dont take part in color. Color actually proceeds from the lamellae that
are excessively long among certain structural scalesproducing real multilayers, sometimes up to a dozen!
Other structures can be contained in the scale thickness, thus not visible by using SEM.

A certain amount of scales are provided with sponds to about a quarter of the visible wave-
more or less complex and regular structures lengths of the spectrum and leads to colored
on their external membrane or into the mem- optical effects. These scales are designated as
brane itself. The standard dimension of those structural. Others, which dont tend to show
structures is comprised between a hundred such structures, have stocked a certain quantity
and several hundred nanometers, which corre- of pigments, either in the constitutive material

Figure 3.34. Scheme representing the arrangement of lamellae in striae. Lamellae are in equal number
anywhere on the stria and form, with the scale superior membrane, an angle that can vary by around 10
according to species.
34 Chapter 3 Lepidoptera Description and Scales of Observation

(a) (b)

(c) (d)

Figure 3.35. Diverse structures of striae. SEM view of Morpho godarti structuralabove, on the leftand
pigmentary striaeabove, right. Structural striae here consist of ve overlapping lamellae kept wide apart
from the membrane by a system of long arches. Lamellae, their overall covered surface attened, are almost
cylindrical at their ends. They are linked to one another by a thin network of vertical ridges around a hundred
nanometers apart. Pigmentary striae actually present the same structure, yet lamellae, very short, do not
overlap. The widening main arches are present all along the inter-striae space, forming small compartments,
in which granules of pigment can be found among some species. Reinforcement pieces are well developed.
Below, Archeoprepona striae: pigmentary on the right and structural on the left. The general arrangement
is roughly similar to that of Morphos, but the structural scales lamellae, which are cover scales here, are
peculiarly distorted.

itself or externally as granulesthese scales tary scales are often lobed and overlap on a
are pigmentary. The distinction is not always signicant part of their surface.
so denite though, and numerous structural Among scales with a well-established role,
scales can also include diffused pigments in we can mention the beautiful androconia,
their structures. a characteristic of Lepidoptera. Aphrodisiac,
Structural scales generally show simple these scales are often very long, sometimes
geometrical shapes that can efciently re- ending in various capitate laments or plumes.
cover without overlapping, whereas pigmen- They are hollow and relate to a glandular cell
Description of the Wings 35

(a) (b)

(c) (d)

(e)

Figure 3.36. Various inter-striae spacings. From top to bottom and left to right, two views of Polyommatus
icarus, and Papilio machaon, Morpho cypris, and Catopsilia orella.
36 Chapter 3 Lepidoptera Description and Scales of Observation

Figure 3.37. Multilayer structure of a


Procris statice scale observed by SEM.
The cross section of a scale cut through
the middle reveals the main trabec-
ulae joining the two membranes of
the scales and the superior mem-
brane multilayer structure. The struc-
ture consists in six very irregular chiti-
neous layers that are alternated with
constant spacing with a thick network
of microtrabeculae. The exterior layer
shows numerous alveolae. The pene-
tration speed of diverse liquids in the
structure proves to be high, suggest-
ing that the other layers must contain
alveolae as well.

the scented secretion of which goes through the covering the latter by overlapping each other
cuticle and spreads its fragrance around the in- a little bit. Their shape is usually simple with
sect. Androconia can be either grouped in var- regular or slightly dense edges.
ious spots on the wing, or scattered all over the As concerns physical colors, scales are often
surface between the cover scales. arranged in two layers: the under layer scales,
known as ground scales and the outer ones, or
Arrangement of the Scales cover scales on top, which are always very dif-
There is no characteristic arrangement of scales ferent in terms of shapes and structures. De-
on the membrane. If in primitive groups scales pending on species, structural scales can ei-
are spread about without any apparent logic, ther cover (Archeoprepona, Uranidae . . . ) or line
they canespecially among the developed and at the bottom (Morpho), but in the last case,
vividly colored species that concern usshow the covering scale must be transparent or at-
a quasi crystalline-like arrangement. As we will rophied so that light can reach the structural
see in our physical study of the optical proper- scales. In the rst case, the ground scales are
ties of wings, the alary membrane only takes often highly pigmented and form a dark screen
part in the infrared properties of the wing, and against which other scales stand out. Ground
not in the visible coloration. Whatever their and cover scales emerge alternatively along the
originphysical or pigmentarycolors arise same line, the ones covering the others, ground
in the scales. An isolated scale includes all the ones spreading almost over the whole surface
effects of the whole wing. The way it is im- without overlapping.
planted and arranged among the other scales However, we will mention certain species
may greatly inuence the nal aspect. of Morpho among which this distinction is not
With the exception of scattering, colors of a obvious anymore. While presenting structural
physical origin imply a great regularity of the colors, cover scales eventually slip between the
structure, at the microscopic level as well as at ground scales. Covered with only one layer of
the macroscopic, and the effect could be under- iridescent scales, the buttery is radiant. The
mined, or even annihilated if scales were dis- two types of scales have almost similar shapes.
organized. That is why the structures, shapes, However, if cover scales are composed of struc-
and implantation of iridescent scales show a tures, they still arent pigmentary. When im-
rather strict order. They tend to be parallel mersed in a liquid of index, they disappear
to the membrane or slightly inclined, evenly completely and only ground scales remain.
Description of the Wings 37

Figure 3.38. Several internal structures. From top to bottom and left to right;
(a) Papilio blumeia cover scale on a striae
(b) The same butteryground scales
(c) Ornithoptera priamus poseidona cover scale on top of two ground scales
(d) The sameanother type of cover scales.

This is not the case for colors proceeding Shape of Scales


from a pigment or obtained by scattering. As we have already mentioned, scales present a
Those effects are not very sensitive to the wide range of different shapes and sizes, in re-
incidence of light and dont require any lation to their roles (camouage, aerodynamic,
specic arrangement. Scales can be more or sexual, thermal regulation, hydrophoby . . . ).
less inclined on the membrane and show a The smaller ones measure less than a few
great variety of shapes. microns, whereas the larger ones are several
38 Chapter 3 Lepidoptera Description and Scales of Observation

hundred microns long. Either pigmentary or striae, in most cases parallel to their main
structural, almost all of them present on their axismore rarely oblique or as a spiralmore
superior side a pattern of longitudinal striae or less thick, spaced, and easily distinguish-
and also some other transversal structures, able thanks to an optical microscope. Over the
which we will describe in detail later. Here largest surface of the scales, striae tend to be
is the source of colors, which can still be parallel to each other, crossing the scale from
greatly affected by the general shape of the one side to the other and regularly spaced,
scales. The pigmentary and physical effects especially on the scales producing physical
are clearly distinguishable on at or slightly colors. On the contrary, they can be totally
curved scales. Structures are always situated disorganized on some specialized scales like
on the superior face, while potential pigments androconia.
are found on the inferior face or among the The distance separating each main striae
ground scales. The whole structural surface is varies from 500 nm on certain structural
exposed to radiance. But numerous scales in- scales, which corresponds to 2000 striae per
clude more complex sections (concave, con- mm, to more than 2 m on pigmentary
vex, undulating, or twisted), which can lead scales.
to combined effects that resemble iridescence, The thickness of striae also varies substan-
without being iridescence stricto sensu. This is tially. The highest ones can be as high as a
the case of certain concave structural scales, micron (among Morphos for instance), but less
the apex of which is directed towards the ex- than the hundred of nanometers. All striae
terior of the wing, thus revealing the infe- share a rather simple structure, which each
rior pigmented face of the scales, when ob- species will modify at will by developing or
served from the apex towards the base. In inhibiting certain elements, sometimes exces-
this direction, they look matte and brown, sively.
whereas they turn iridescent in the opposite di- Striae are composed of more or less long
rection. According to this principle, scales pre- lamellae, arising at regular intervals on the
senting several undulations show from a dis- superior membrane of the scale and overlap-
tance successive bright stripes corresponding ping like tiles. Lamellae are thus all inclined
to the crests of the waves and dark stripes cor- on the scale according to an angle of about
responding to the troughs, and thus create a 10 . The longer the lamellae are, the more
changing combination thanks to the incidence they are to overlap. Hence, there are some
of viewpoint. However, this is more the result cases where as much as 12 lamellae overlap.
of effects of geometry and shade than strictly Their length is also remarkably constant and
iridescence. almost equal to a whole multiple of the in-
The morphology of the scales has a great in- terval between their bases, so that the num-
uence on the color of the wing, as well as its ber of layers is constant on every point of the
intensity, purity, and on optical effects in gen- stria.
eral. And yet, the diversity of geometries is
such that it is impossible even to mention them.
The Interstriae Spacing
We will study in detail a few cases to convey
this diversity. The diversity is as great concerning interstriae
spacing. The latter plays a signicant role espe-
cially among butteries that diffuse light since
Fine Structures; Striae and
it is there that structures take place. We will
Interstriae Spacing
describe them fully in the chapter dedicated
We have already mentioned the complex sys- to scattering and will now only give an idea
tem of striae present on the superior surface of of their diversity. Let us mention that this area
the scales. This characteristic is quasi general in the membrane often shows pores, which al-
among Lepidoptera. Apart from a few rare low the gas exchanges with the interior of the
specic scales, all of them include longitudinal scales.
Description of the Wings 39

Internal Structures of Scales Figure 3.28, examined with a transmission elec-


tron microscope. In a few rather simple cases,
Always on the same scale, one can notice struc-
the structures that are optically active are sit-
tures not in striae anymore, but in the supe-
uated in the membrane, as we have just seen
rior membrane of the scale itself, for instance,
with the Uranidae. However, these structures
among a whole family of Lepidoptera such as
may be situated in the striae themselves and
the Urania. The membrane is thus not com-
not in the supporting membrane. This arrange-
posed of a single chitin layer anymore, but
ment resembles that of Morphos, yet here the
consists of overlapping layers, sometimes as
stria doesnt form a surface structure by trans-
many as six or seven, each measuring about
forming itself, but it creates an internal archi-
a hundred nanometers and separated at con-
tecture.
stant intervals by microtrabeculae. These lay-
The two arrangements can be combined.
ers often contain numerous alveolae enabling
Lastly, the membrane may also consist of a
gas or liquids to circulate in the system. The
structure composed of alveolae, capable of pro-
internal structures of the scales are extremely
ducing colors through scattering.
diverse, as illustrated by the series of sections in
4
Coleoptera Description and
Observation Scales

Coleoptera represents an extraordinary order caterpillars, have legs and pseudopods, while
from every point of view and we will study it oligopod ones are not provided with pseu-
in detail. It is indeed the most important or- dopods, and apodes larvae at least, which
der in the animal world, including the biggest are even completely deprived of any locomo-
insects but especially the most colorful ones. tive organs. Most of Coleoptera have oligo-
Coleoptera appeared before Lepidoptera some pod (carabidae, scarabaeidae) or apod (scolyti-
280 million years ago. They are characterized dae) larvae. These larvae undergo successive
by sclerotic anterior wings: the elytrons. The metamorphoses before reaching the nymphal
latter are not involved in ying but form an ex- stage.
tremely strong protective case (koleos in Greek) If numerous Coleoptera are carnivorous
after which the group was named. and therefore possess strong mandibles, many
have also developed some chemical protec-
tion, which make them repellent in terms
Coleoptera of taste or often even toxic. Like butter-
ies, Coleoptera use Batesian and Mullerian
As indicated by their name, Coleoptera are mimicry, in which their extraordinary colors
characterized by elytrons, anterior sclerotic play a critical part, of course. We wont elab-
wings that recover and protect the abdomen. orate on the subject.
Posterior wings are the only ones used for y- These colors generally recover the whole
ing; they are membranous and folded both lon- body, including legs, antennas . . . and they are
gitudinally and transversally under elytrons, particularly visible among elytrons. Among
which remain folded or half-open during Coleoptera, the cuticle is an incredibly com-
ying. plex material that we will study later. Let us
Among many Coleoptera that lost the ca- notice for now that, contrary to butteries and
pacity of ying, posterior wings can be partly even concerning physical colors, Coleoptera
or totally atrophied. The order is divided into can present optical structures in their scales,
these suborders: archostemata, myxophaga, inside of the wing itself and sometimes in both
and adephaga, composed of the carnivo- at the same time. This must be the price for the
rous Coleoptera including cicindelidae and most beautiful colors existing in nature.
polyphaga. They can be found in almost ev-
ery environment, including running and sea
The Cuticle of Coleoptera
waters.
The great variety of insect larvae is usu- In order to address this second place where
ally classied according to functional rather color can be found that is the carapace
than phylogenetic criteria; hence the tradi- of Coleoptera, we must describe this ex-
tional distinction between polypod, oligopod, traordinary structure covering arthropods: the
and apod larvae. The polypod larvae, like the cuticle.

40
Coleoptera 41

Figure 4.4. Schematic representation of a ying


Coleopteron. The general outline is similar to that
of butteries. Anterior wings, which are very atro-
Figure 4.1. Campsosternus head. Colors are interfer- phied and after which the group was named, remain
ential. half open or folded during ight and kept together
by a coopting device. Venation is roughly similar yet
more complex than among Lepidopterons.

This highly complicated integument plays


a major role in the success of insects. It both
supports and protects the insect, providing
muscles with bases and regulating liquid and
gas exchanges with the environment. It is also
the center of the magnicent colors that some
Coleoptera sometimes show, making them
Figure 4.2. Diverse Coleopterons larvae. (a) Scolyti-
dae apode larva. (b) Scarabaediae ligopod larva look like exquisite pieces of jewelry. This last
(according to Chu, 1949). aspect particularly interests us, that is why
we will examine this structure in detail. Like
among Lepidoptera, this structure is some-
times covered by scales that can totally or par-
tially generate colors themselves or modify
those produced by the supporting structure.
Most of the scales of Coleoptera are thick
and without surface structures. However, oth-
ers are covered in protuberances measuring
some hundreds of nanometers that diffuse
light and break the shine of the scales. Scales
broken after freezing show an almost perfect
crystalline structure, which as we will see, orig-
inates diffractive phenomena resulting in ex-
traordinary effects of color.
Structure of the Cuticle
The cuticle is a complex system of various lay-
ers in which one can distinguish two main
parts: a thin external layer known as epicuticle
Figure 4.3. The life cycle of a Coleopteron with
apode larva. The latter undergoes several sheddings and a thicker internal one covering the epider-
before the nal nymphal stage (according to Gullan mis, called the procuticle. The two big layers,
et al., 2000). and especially the rst one, are also divided
42 Chapter 4 Coleoptera Description and Observation Scales

Figure 4.5. Chitin at vari-


ous magnications. (a) Two
monomers of the molecule
(b) The microbrils arrange-
ments of molecule groups.

(a) (b)

into many sub-layers that can individually play and oriented parallel to the epicuticle, forming
a part in the coloration of the elytron thanks to layers that produce lamellae by superposing.
their structures. It is this system of lamellaes that originates the
We wont elaborate on this subdivision, the numerous optical phenomena resulting in the
classication of which hasnt been established physical colors of Coleoptera.
yet, but we will study one of the components At this point, helicoid arrangement can oc-
that play an essential role in the elytron color: cur. This is a most peculiar and characteris-
chitin. tic phenomenon providing arthropods with
The chitin molecule is a polysaccharide unique properties of circular polarization.
with a structure similar to vegetal cellulose.
It is a long chain composed of monomers of Helicoidal Structure
N-acetylglucosamine. About twenty in total, In the exocuticle, layers are parallel to each
these molecules form two or three rows, thus other, whereas each microbril turns following
creating almost crystalline sticks: the microb- a relatively constant angle, which results in a
rils. The latter are coated with a proteinic matrix helicoid structure similar to that of some liquid

Epicuticle
Exocuticle

Cement

Waxe
Procuticle

Superficial layer
Cuticle

Outer epicuticle
Endocuticle

Inner epicuticle

Exocuticle
Figure 4.6. The main subdi-
visions of arthropodes cuti-
cle. Divided between endo-
cuticle and exocuticle, the
procuticle lies on the epider-
Epidermis mis cells and is covered by a
very ne epicuticle. Numer-
ous ducts and canaliculi run
across the whole.
Coleoptera 43

Figure 4.7. SEM image of an


helicodal structure in the cas-
side cuticle. The elytron un-
derwent an optical polishing
at a circular alveola, followed
by an ion etching revealing the
orientations of leaves in each
layer.

crystals in the cholesteric phase. Yves Bouli- The analogy with the cholesteric phase of liq-
gand is the rst who demonstrated it concern- uid crystals may be fortuitous, yet the formal-
ing seashells and later, Neville demonstrated it ism elaborated by Fridel will allow us to briey
concerning insects. The oblique section of such study the phenomenon.
a structure shows a puzzling arrangement of This phase appears with molecules known as
arches, a result of this helicoid piling accord- chirals, which means that their image cannot be
ing to Bouligand. superposed in a mirror. This is the case of every
One can also notice it on a section of the tu- organic molecule including at least one asym-
bercles or concavities in the elytron. The lay- metrical carbon with four different groups.
ers thus look like concentric circular bands The main physical result of this characteristic,
in which brilles are all parallel to each called chirality, is an optical activity, which en-
other. tails that the direction of the light polarization

Figure 4.9. In a cholesteric stage as in the cuticle


Figure 4.8. A cross section through the endocu- in Coleopterons, the director orientation regularly
ticle points out arched structures characterizing varies along the axis. A period shows up when the el-
cholesteric-type arrangement. Lamellae are formed ements underwent a 180 rotation, determining thus
by a 180 rotation of sticks. a lamella.
44 Chapter 4 Coleoptera Description and Observation Scales

with phenomena of polarization by reection


caused by the carvings and alveoles sometimes
adorning the surface.

Cuticle and Scales


The color of Coleoptera originates from the cu-
ticle or, as we have seen, from the scales cover-
ing it. The structures that can be found are very
different and produce various colored effects,
which cover a larger part of the spectrum than
the colors of butteries. We will overview those
Figure 4.10. When entering the cuticle, the polariza- different structures.
tion direction of a polarized wave rotates according
When color comes from a structure included
to the medium director.
in the cuticle, the surface itself is only involved
in the spatial scattering of light. One can nd
rotates when crossing such a medium, with the almost any kind of surface structures, from the
associated interference phenomena. perfectly smooth surfaces of cetonia aurata and
In each plane and because of their elongated other scarabs to the alveolar surfaces of cin-
shape, molecules present a well-specied ori- cidelidae, or hillocky ones of cassidae, or even
entation, called director. In the cholesteric phase longhorn beetles . . . In almost every case, color
like in the endocuticle, the orientation varies is interferential but the colored effect is mod-
from one plane to the next one, constantly re- ied by the surface structure, which can scat-
volving around an axis perpendicular to the ter color and thus decrease its brightness, on
molecules. the one hand, and mix the different shades,
This rotation is known as twist. Among in- thus undermining the spectral purity, on the
sects, it goes clockwise from the viewpoint other hand. These surface structures show di-
of an observer looking at a point. The char- verse sizes, ranging from a millimeter among
acteristic length on which the director un- cassidae for instance . . . to about ten microme-
dergoes a rotation of 2 is called period of ters among cincidelidae.
the cholesteric and forms two lamellaes in the In any case, the hierarchy of the scales of
cuticle. sizes and colored effects of Coleoptera is quite
Every layer actually behaves like a uniax- similar to that of butteries. Like among cin-
ial anisotropic medium with a slow axis par- cidelidae as represented on the next page,
allel to molecules and a quick perpendicular the elytron, which is rather dull, possesses
axis. These two axes regularly revolve layer by an alveolar structure. When examined with
layer when light enters the cuticle, which then a photonic microscope, the alveoles, measur-
behaves, if the step is of an appropriate size, ing about ten microns, look extremely color-
as a multilayer generating interferential colors. ful with chroma covering the whole spectrum,
This rotation of the index also gives an optical from the red to the blue with green as a pre-
activity to the whole, making the direction of dominant color. This proceeds from a pointillist
polarization of the wave revolve in the same effect. A cross-section reveals the multi-layer
way and with the same step. structure that produces the color. We will see
The epicuticle is also overlapped to this al- in the chapter on interferences that the basin
ready quite complex structure. The former can shape introduces, on a smaller scale, a chro-
include a single-multilayer structure consisting matic component and generate polarization
of isotropic lamellaesin which no brillary effects.
structure has been established. It sometimes Among many species on the contrary, the
also originates interferences, often combined cuticle doesnt produce any physical color but
Coleoptera 45

(a) (b)

(c)

Figure 4.11. Views of the surface structures of elytrons at different scales. Millimetric alveolae in a cassida:
below. In the center, the slightly depolished surface of a Scarabeide Eudicella graf. Left, Plusiotis cupreomaig-
inata smooth surface.
46 Chapter 4 Coleoptera Description and Observation Scales

Figure 4.12. ColeopteronCincidela


campestrisat different magnica-
tions, from the elytron to the external
epicuticle.

Figure 4.13. The lined scales of Hoplia


argentea.
Coleoptera 47

Figure 4.14. Coleopteron with


scalesCurculionidaeat dif-
ferent magnications. The bulk
scales (massive) only partially
cover the elytron. A broken
scale reveals the crystalline
structure, the base of diffrac-
tion phenomena generating
colors.

(a) (b)

Figure 4.15. Various arrangements of Coleopteron scales. Above, sh-type arrangement: Hoplia cerulea C.
Scales overlap to a large extent. Below, scales are more scattered, like pebbles: Curculionidae.
48 Chapter 4 Coleoptera Description and Observation Scales

is covered with scales like Lepidoptera. Their cuticle. However, there are cases presenting a
structure is nevertheless quite different since sh type arrangement in which thin and at
we always nd the same internal crystalline scales overlap.
organization, which is rare among butteries. With a few exceptions of the latter kind, the
There are numerous families of Coleoptera pro- scales of Coleoptera are thick, hard, circular,
vided with colored scales. The principal one is and convex. They are usually attened against
the Curculionidae, which includes our dear and the elytron, making it difcult to perceive the
rather dull weevils, but also the Scarabeidae peduncle. Their external surface is the more
and especially cetonia, well known for their ex- often loosely folded. However, it can be cov-
traordinary metallic shine. ered with tiny nipples resembling those of the
Contrary to Lepidoptera, scales are gener- corneal lens of ommatidium (compound eyes
ally arranged without apparent order and often of many butteries), where they can play the
scattered, discovering most of the underlying same role of antireecting layer.
5
Changing Colors: Structures
or Pigments?

A Few Basic Experiments which shows a surprising metallic aspect, the


pigment of which has recently been isolated.
If, as we have seen in the introduction, the pig- The electron microscope reveals an arrange-
mentary or structural origin of the colors of ment of rather similar scales relatively raised
Lepidoptera was established a long ago, the at the apex, which contributes to the result-
classication of a given buttery in one of the ing iridescence strictly due to a geometric
two categories has not been as easy and re- effect.
mains blurry today. This is generally due to The scales themselves present no structure
the combination of structural and pigmentary capable of producing such brightness. The
effects, but also to the multiplicity of structural wonderful Ornithoptera priamus poseidon offers
effects. Let us make a list of the phenomena a similar case. Its bright yellow glint apparently
as the biologist H. Onslow, for instance, would does not result from any structure of a surface
present them in the early century. or volume.
Concerning structures, the phenomena of
diffraction, refraction, interferences, and scat-
Pigments tering were commonly distinguished, whereas
they are not fundamentally different and it is
Like those used by painters, pigments work by not easy to separate them. All proceed from the
selectively absorbing a certain range of wave- diffraction of an electromagnetic wave by an
lengths. Observed in transmissionwhich is obstacle. Their difference is more due to the ge-
rare because they are usually opaquethe pig- ometry and arrangement of the obstacles than
mented objects show the complementary color to physics itself. It is nevertheless more conve-
of the dominant that has been absorbed. This nient to rst study them separately, even if we
becomes more complicated with reection as will later mention them all together.
we will see, since it involves the phenomenon In order to explain the metallic and iridescent
of selective reection, which scarcely leads to colors among certain insects, four categories
pigmentary coloration. Fluorescence phenom- have been established:
ena also belong to this category.
On a regular basis, the absorption peaks of r The diffraction of the wave by a grating type
the pigments are wide, resulting in not very periodic structure
pure colors. Furthermore, they are often as- r Light interferences on the surface of thin
sociated with scattering structures that break single-layer or multilayer lms
specular reections, making them look rather r Light refraction caused by prismatic struc-
matte. There are many exceptions in which tures
very bright colors derive from a pigment, as r The selective scattering of low (blue) wave-
has been established. This is the case of Ly- length by particles or cavities smaller than the
caena phlaeas, the small copper of our mountains, wavelength

49
50 Chapter 5 Changing Colors: Structures or Pigments?

that of the material constituting the structure.


When pigments are dissolved, the color van-
ishes irreversibly, and one must check that the
scale structure hasnt been altered during this
rather aggressive process. When the origin is
structural, the process is reversible. As the liq-
uid evaporates, the wing retrieves its initial
color. The process can also be partial; if the in-
dex of the liquid is different from that of the
structure, the color is modied. This implies
that the structure is open and somehow lets the
liquid penetrate it and then get out. We will
mention the possible consequences of such a
Figure 5.1. Iridescence effects in Lycaena phlaeas, il- treatment on a pigmented surface in our study
lumination coming from the right. of pigments.
We will illustrate our point with the
In the following chapters, we will rst focus 2-dimensional structures of a Morpho type.
on each of these phenomena by studying spe- Their big size offers a better support for
cic buttery wings that show one of these phe- study and they show the most spectacular ef-
nomena predominantly. However, such phe- fects. Besides, the structure of iridescent scales
nomenon seems to be always combined with among Morphos is very open, allowing index
others, which is why we will then mention the liquids to rapidly circulate. We will see in the
combinations of the phenomena, which will chapter on interferential phenomena that for a
ruin our illusory yet convenient distinctions. given conguration, a minimum occurs in the
It is in many instances convenient to distin- reectance, hence the extinction of the corre-
guish between pigmentary and structural col- sponding color, for each wavelength such that:
ors. There are two methods to do so: either re-
moving the pigment by dipping the wings into kmin = 2ne cos r (5-1)
various solvents, or removing the structure
from the optical point of viewby dipping it where n is the layer index, e its thick-
into a liquid of optical index exactly equal to ness, k an integer, and r the refraction angle.

Figure 5.2. SEM view of Ly-


caena phlaeas scales. Cover and
ground scales show differ-
ent lengths, but similar thin
structures.
Pigments 51

changing one of these parameters. Let us be-


gin with the index variations.
An increase in the index of one of the com-
ponents of the multilayer obtained by dipping
an insect into a liquid with an increasing index
leads to a shift of the dominant color towards
the high wavelengths (blue green yellow
red) but also to a decrease of the reectance
(or increase of the transmittance) resulting in
the disappearance of red colors. When indices
are equal, interference and thus color disappear
and the wing becomes transparent. If such a
structure is laid on a pigmentary motif or back-
ground, those only will be visible.
Other elements can also help us. The modi-
cation of colors under various index liquids cer-
tainly provides the most convincing test. How-
ever, it is little selective and doesnt allow one
to distinguish between the different structural
origins of colors. As thin layer interferences are
concerned, the position of reection maxima
and thus the resulting color not only depends
on the refraction index of the various layers,
but also on their thickness. The same effects
can be observed (yet they are more difcult to
illustrate on a whole buttery) by changing the
pressure that structures are subjected to. Until
the physical destruction of the latter, the effect
is reversible and can easily be examined with
an optical microscope.
Finally, it is possible to separate the pigment
and the structure effects by making the opti-
cal measurements on an impress. However, we
will see that some structural effects occur not on
the surface but within the scale and thus cannot
be revealed by this technique. In other cases, it
allows us to distinguish between interference
and diffraction effects. From a strictly optical
point of view, the fact that colors proceed from
a structure and not a pigment has important
consequences. In particular, as there is no ab-
sorption, the incident ux is shared between
Figure 5.3. Colors variations in Morpho menelaus, the reection and transmission. In the visible, the
scales of which are highly pigmented. When the color that is transmitted is almost the comple-
matching of indices is reached, it turns almost black. mentary of the reected color. In reality, this
is not easy to observe because of the constant
Under a given incidence, the reected wave- presence of pigments, even in small quantity,
length, and so the observed color, depends on within the scales or membrane structure.
the product of the layer thickness and with We wont go beyond this simple explana-
its optical index, and can therefore vary by tion. The variations in certain colors of buttery
52 Chapter 5 Changing Colors: Structures or Pigments?

Figure 5.4. Colors variations in Morpho godarti under


different index liquids. On the right, the butter-
y in its normal state in the open air. In the cen-
ter, immersed in acetone n = 1.362and in
trichloretylen on the left hand-siden = 1.478. Mor-
pho godarti structural scales hardly contain any pig-
ment and turn transparent when the index matching
is reached, thus exposing the ventral side motifs.

wings can be explained with the index of the


medium by the traditional theories of geo- Figure 5.5. Indonesian Ebonia9 cm. The anterior
wings of this buttery, which belongs to the Pieri-
metric and physical optics: scattering, inter-
dae family, are totally pigmented in orange. Posterior
ferences, or diffraction. We will call these col- wingsyellow-whiteare highly scattering. Under
ors structural. Others proceed from absorption index liquid, they turn transparent, exposing the
and selective reection phenomena. We will ventral side motifs, while anterior ones remain al-
call them pigmentary. It is essential to remem- most unchanged. The pigment is contained in small
granules also scattering light. When not illuminated,
ber that this distinction doesnt hold in real-
wings look darker.
ity. If there are many a buttery colored with
a pigment, I dont know any owing its color
to physics only. The spectacular results always
proceed from a subtle mix between the two
phenomena, the two reinforcing each other, an-
nihilating or combining.

A Few General Principles


Based on this rst remark, we will examine
each of these physical phenomena more closely
as presented by specic butteries. Although,
they dont establish strict rules and the reader
Figure 5.6. Spherical coordinates system determin-
is invited to nd the counter-example, here ing the source position. The buttery is observed ac-
are a few principles that can help us in our cording to the Oy axis normal to the dorsal-ventral
choice. plane of the insect.
A Few General Principles 53

Figure 5.7. The big Indonesian


Papilio, Papilio ulysse, illuminated
under two extreme angles. On
the right, under low-angle from
the left. The predominant wave-
length is in the blue-violet range,
around 475 nm, approaching pur-
ples in places. On the left, the
insect is illuminated and ob-
served under normal incidence.
Its color tends toward the green
500 nm. In both cases, there
are no signicant differences be-
tween the left and right parts of
the insect, suggesting there is a
symmetry in the incident plane.

Let us begin with basic remarks: Pigmentary colors or colors proceeding from
selective scattering hardly vary, whatever
Blue pigments are rare in the animal world.
the angle of observation or illumination, ex-
Most of them are metallo-protein pigments
cept when angles are very high and for
characterized by the presence of one or sev-
highly polarized light. The corresponding
eral metals on various protein molecules,
aspect is matte, even dull.
the most common of which is haemocyanin,
which derives from copper. It is usually These experimental techniquesmodica-
found among mollusks, crustaceans, and tion of the index or of thicknesspermit us
some arachnids in solution coloring plasma in many cases to distinguish between struc-
blue. tural and pigmentary origin of colors. They
On the contrary, red, yellow and brown pig- always lead to the alteration or disappear-
ments are widespread, especially among ance of colorsa real opportunity for nature
butteries, either as a product of degradation and the opportunity for us to mention all the
(such as melanins, widespread in the animal types of color alterations, of chroma, observed
world), or as tetrapyrolique pigments (like among insects. Let us only note that color al-
hemoglobin) from which derive the vivid terations are quite rare among the latter. There
reds of numerous vanessas (Inachis io, Aglas are physiological modications, rather quick
urticae . . . ). and generally reversible, and morphological
changes generally irreversible resulting in a
If we now examine the wings from the point
new pigment. We will focus on the former,
of view of the variations in colors related to
which are obtained through the modication
the angle of incidence, we can establish a new
of the cuticle structure or the migration of
selection:
pigments.
Interferential or diffractive colors change sub- We will call the devices leading to color vari-
stantially depending on the angle of observa- ation with the generic term X-chrome, X desig-
tion or illumination. This is the phenomenon nating the factor causing the variation, thermo
of iridescence, one of the most spectacu- for temperature, gonio for the angle, and hy-
lar among tropical butteries like Morphos, gro for humidity . . . Many of those phenom-
Archeoprepona, and Urania, but also, for in- ena dont involve color strictly speaking but
stance, among the little common forester rather brightness, that is why we will briey
(Procris Statice) in temperate regions. The mention them. Goniochromic structures on the
aspect of iridescence is here bright and contrary are the very focus of the present
metallic. book.
54 Chapter 5 Changing Colors: Structures or Pigments?

Figure 5.8. The viewpoint is different here, but the


light source remains unchanged, here in the equato-
rial plane at 60 on the right to the normal. The ob-
server is in the same plane, at 30 to the normal, on its
right or left. Color varies even more; from yellow
560 nmit turns blue 496 nm. Colors also
appear in the basal black pigmentary areas and in
the exterior edges of anterior wings. Papilio blumei
10cm.

Changing Colors
Goniochromic Structures
They strictly concern iridescence. Indeed, color
changes according to the viewpoint or angle
(gonio in Greek) of light, which means that color
is physical here, except for the very specic

Figure 5.10. (a) The yellow form of Dynastes hercules.


(b) In a humid atmosphere, the naturally diffusing
external cuticle gets soaked with water and becomes
more transparent. Light is absorbed by deep layers.

cases of selective reection already mentioned.


The angles of observation and illumination are
Figure 5.9. Active color change in the Australian
Cassida, Notasacanta dorsalis. Under stress, a liquid generally established within a system of spher-
penetrates the structure and increases the optical ical coordinates. Those effects are often spec-
thickness of permeable layers. tacular and accompanied with polarization
Changing Colors 55

ors among insects. These variations are not


related to the ambient humidity nor to the
alternation of night and day, but correspond
to the reaction to an aggression. This device
is commonly used among other classes for
camouage or offense. This is the case of
Cephalopodes (octopus, cuttlesh, and squid),
some skin areas of which present a multi-
layer cytoplasm/water structure, the thickness
of which can be quickly modied by the ani-
mal. The other color changes are strictly pas-
sive as they depend on exterior physical condi-
tions (temperature, hygrometry . . . ) and they
more often lead to a change in the thermo-
dynamic potential of the animal. In this case,
there is no variations in color but in brightness.
Among the South American big Scarabaeidae
Dynastes hercules, the external cuticle is com-
posed of a yellow spongy layer covered by a
Figure 5.11. Thermic regulation in Koscuiscola. Un- thinner transparent layer.
der high temperature, small inclusions are on the In the sun, the animal looks yellow-brown.
surface and scatter light. When temperature is lower,
a mix occurs and absorbs all wavelengths, the insect But when hygrometry increases, the spongy
turning black. layer becomes impregnated with water and
turns transparent, thus letting light penetrate
phenomena. We will further study them in the the deep layers where it is absorbed. The insect
chapters on physical colors. turns matte black. In the forest where those
insects live, such hygrometry variations occur
on a daily basis. Therefore, the insect looks
Thermochromic, Hydrochromic, and Other
brown during the day when it feeds under the
Structures
trees and black at night, becoming invisible to
We have mentioned that when the optical in- its predators.
dex of one of a components in a multilayer Another modication due to thermochromi-
was changed, it could change the color of ism: the darkening of the grasshopper Koscius-
a Morphos from blue to green and nally to cola as well as many Odonata and Phasmida.
brown. This cannot occur in the reality. The During the day when the temperature is mild,
only uid a buttery could use, waterthe in- Kosciuscola looks bluish due to the scatter-
dex of which n 1.35 would be perfect for such ing of the shortest wavelengths by small in-
a phenomenonis yet not wet enough in or- clusions consisting of leucopterines and uric
der to penetrate the thin structures of scales. acid (less than 0,2 m high) on the surface of
However, there is at least one Australian cas- the integument. At night, when the temper-
sida, Notasacanta dorsalis, that uses this tech- ature is falling, bigger pigmentary inclusions
nique. The colors are created in a multilayer, come up to the surface and absorb all visi-
the hygrometry of which can be changed by ble wavelengths. And the insect looks black. It
the insect, resulting in the variation in the in- is assumed that these thermo-chromatic vari-
dex and thickness of layers. This is the only ations allow insects to regulate their tempera-
known occurrence of an active change of col- ture more rapidly.
6
Physical Colors, Chemical Colors
Basics of Solid State Optics

LightMatter Interaction, on to the different phenomena that light pro-


Polarization, and Optical Index duces in the material, which are character-
ized by the index. Under its wave aspect, a
Now that we have completed our tour of but- ray of light consists of the propagation of two
tery colors and of their advantages for a elds: an electric eld E, and a magnetic eld
given species and mentioned how difcult it H. They are perpendicular to each other and
could be to interpret them in the past, we will to the direction of propagation. As a conse-
examine their origin in detail and try to of- quence, the electromagnetic wave is transverse.
fer a simple view of todays knowledge on The spatiotemporal dependencies of the elds
this matter. We have already mentioned op- were established in the 19th century and ex-
tical index in the last chapter, and a few ba- pressed by J. C. Maxwell in his famous equa-
sic experiments in which butteries are im- tions that were named after him. We wont
mersed in various liquids. These experiments present these equations but only one of their
and their effects on colors are the keys for simplest solutionsthe harmonic plane wave
solving the problem. When one knows the (Figure 6.1) and a more complex solution that is
optical index and the geometry of a given ma- critical for the study of Coleopterawave with
terial, one automatically knows its color. Con- circular polarization. Within such waves, E and
trary to Raleigh and Michelson, we now have H elds vibrate in the same plane, called wave
access to the geometry thanks to electronic plane, and propagate perpendicularly to this
scanning technologiesscanning and trans- plane.
mission electronic microscopesand atomic In natural light, like sunlight, no direction of
force microscopes. Yet, what about the optical vibration is imposedthe wave is said to be
index? nonpolarized. Always perpendicular to each
It would go well beyond our subject to men- other, E and H elds can vibrate in any direc-
tion all the aspects of optical index theory con- tion of the plane. It is yet possible to force the
cerning the various kinds of materials. This elds to take a specic direction of vibration
would leave butteries aside and moreover, or to change a preexisting direction by mak-
materials that concern us are all of the same ing the wave interact with polarizers. Polaroid
type. They are insulating organic materials, so lms offer one of the most common examples
it would be pointless to study metals or semi- of polarizers. If they are not as effective as the
conductors. We will therefore focus on general latter, Coleoptera and butteries are also often
principles necessary to understand the physi- sources of polarization. It is one of their most
cal phenomena that will be described. The in- remarkable properties. Before studying further
dex is the parameter governing the interac- this aspect, let us nish presenting the mag-
tion of light with matter. We will rst briey netic eld. Since E and H elds are not inde-
describe the former and we will then move pendent, it is pointless to study them both. For

56
LightMatter Interaction, Polarization, and Optical Index 57

Denitions Relations

E = electric eld Frequency: = c/


H = magnetic eld Wave vector: |k| = 2/
k = wave vector Period: T = 1/
W.P. = wave plane
= wavelength

both historical and physiological reasonsthe


eye and most of collectors are sensitive to the
Figure 6.2. The electric elds of a nonpolarized
only electric eldoptical demonstrations tra-
wave falling on a diopter. According to Descartes
ditionally use the latter. H variations can then law, the angle of incident i is equal to the angle of
be deduced from those of E thanks to Maxwells reection r . The electric eld of s wave Es , perpen-
equations. dicular to the plane of incidence P.I., remains in the
diopter plane, whatever the incidence, contrary to
the eld Ep .

Linear Polarization
Let us rst note that when a wave, even non- senkrecht) and the other Ep being parallel to the
polarized, falls on a surface, specic directions plane (parallel, in German).
show up ipso facto. If one calls plane of inci- The two elds dene two perpendicularly
dence the plane dened by emergent and in- polarized waves, which, if they follow Snell-
cident rays, whatever the direction of the in- Descartes laws establishing their propagation
cident eld, it is always possible to express it directions, can interact with matter in differ-
as two vectors: one Es being perpendicular to ent ways. One of the most noticeable con-
the plane of incidence(from the German term sequences is the angle dependence of their

Figure 6.1. According to Maxwell equations, the E


and H elds of an electromagnetic wave are per-
pendicular to each other and to the k propagation
direction. Under nonpolarized natural light, they Figure 6.3. Variations in the reection factors of s
can oscillate in any direction in the wave plane and p waves with the incident angle. For a given
W.P. Under polarized light, a vibration direction was angle of incidence b , known as Brewster angle, the
givenOx for the electric eld and Oy for the mag- reection factor of the p wave is equal to zero; the
netic eld, on the gure above. reected wave is totally s-polarized.
58 Chapter 6 Physical Colors, Chemical Colors Basics of Solid State Optics

Figure 6.4. Decomposition of a wave


linearly polarized into two waves in
phase, Ex and Ey, before crossing any
anisotropic material. In a plane perpen-
dicular to the propagation direction, the
direction of the electric eld E remains
unchanged during its propagation.

reected intensity (Figure 6.3), which is often Coleoptera, that we should present: circular, or
observed among butteries. more generally, elliptic polarization.
Although that of the s wave, with its electric The vector nature of the electromagnetic eld
eld remaining in the layer plane, whatever the entails a matrix treatment of this kind of prob-
incidence, hardly depends on the latter, that lems. We wont try to study this matrix formal-
of the second wave p presents a surprising ism here. We will only mention it through a
angle variation. Under a particular incidence, descriptive and intuitive approach, which will
Brewster incidence, the p wave reectivity is allow us to give a good overview of the differ-
equal to zero! If one illuminates a diopter under ent phenomena observed among insects. When
the B Brewster incidence, the reected light one studies a single vector, like the E eld for
is strictly s polarized. A useful consequence, instance, it is common and convenient to ori-
which is well known among photographers, is ent one of the reference axes parallel to the
that under this incidence, it is possible to sup- vector. Indeed, there is no point in consider-
press the reections of a surface. Another con- ing two componentswhich are not equal to
sequence that directly concerns us is that under zerowhen one is enough.
the same conditions, the light reected by some However, we wont proceed this way. Let
butteries can be highly polarized. It is well us consider a reference in which the vector E
known that contrary to humans, many insects has two components: Ex and Ey (see Figure
are sensitive to light polarization. Their chro- 6.4). In addition, if this eld oscillates with
matic messages are thus enriched with a new a given frequency and moves in a direc-
aspect! The Brewster angle only depends on the tion k, with a speed c, so does each of its
indices of the materials bathing the diopter: components, which thus behave like two inde-
pendent waves and form the two polarization
tgB = n/n0 (6-1) states of the electromagnetic wave. In any wave
Brewster angles of biological materials, with planea plane perpendicular to the propaga-
indices close to n = 1,5, put in air (n0 = 1) range tion directionthe E vector extremity moves
between 50 and 60 . forward and back along a straight line segment.
This proves that the plane wave is indeed po-
Circular Polarization larized linearly.
It is well known that in vacuum, light speed
Linear polarization as mentioned above is a is c 3 108 ms1 . Yet, in any other material,
mere particular state of light, which is im- this speed diminishes proportionally to the op-
portant for us since it causes interesting opti- tical index value of the material:
cal phenomena in butteries. However, there
is another state of light generated by some v = c/n. (6-2)
LightMatter Interaction, Polarization, and Optical Index 59

In a glass block with index 1.5, light speed de- equations:


creases by one third, which is quite signicant. 
X(t) = Ex cos(t x )
And, as will be mentioned later, many beetles
Y(t) = Ey cos(t y )
elytrons consist of a very anisotropic material,
roughly long sticks all oriented in the same di- Where  is the phase difference between the
rection. This provides the elytron with great two components. The vibration direction being
mechanic properties and it also entails that the totally determined by the two equations, they
optical index is not the same for waves vibrat- dene a polarization state. In 1941, R.C. Jones
ing in the stick direction (slow axis) or in the found another expression, with matrix formal-
perpendicular direction (quick axis) in which, ism this time, inspired of quantum mechanics,
according to the formula above, the wave prop- known as Jones vector:
agates more quickly. When going out of such a  
Ex eix
material, the two waves are separated from one V=
Ey eiy
another by a distance d depending on the thick-
ness of the crossed material and on the differ- When light travels through any optical system
ence between the indices in the two directions. S, its polarization state is modied in ampli-
If before crossing the material, the E vector ex- tude, phase . . . and the outgoing Jones vector
tremity of a linearly polarized wave follows a takes another form:
straight line segment as already mentioned, it  
 ix
E e
is not the case at the exit of the medium where V = x

Ey eiy
it in general follows an ellipse. In the specic
case where the two waves have the same am- Jones matrices M express the relations be-
plitude and are shifted from one another of ex- tween the wave polarization states before and
actly one quarter of wavelengththe medium after crossing the optical system.
is then called /4 slidethe ellipse degenerates
into a circle; the wave is circularly polarized! V  = [M]V
We wont further study this phenomenon; wed
rather let Coleoptera illustrate it. Let us only
note that making the wave cross the same /4
slide rotated by an angle of 90 produces the in-
verse process; the two waves are back in phase
and rebuild linearly polarized light. This is how
one can point out circularly polarized waves.
We will illustrate the few most common situ-
Matrix Approach to Polarization ations involving insects with several examples
but without demonstration. We will consider
Our purpose is not to offer a comprehensive the changes only in polarization direction and
study of the diverse polarization states of light not in intensity. All biological media neverthe-
and their transformations. Polarization effects, less present to various extent some absorption,
which are always associated with color effects thus modifying light intensities. Yet, this has
produced by animal structures, represent one little effect on color, so we will neglect this ef-
of the great interests of such a study and gener- fect.
ate many industrial applications. It is therefore Linear polarizer. The simplest case is linear
useful to give an overview of one of the modern polarization, which is produced by glass reec-
approaches of light polarization: Jones matrix tion under Brewster incidence.
representation.  
1 0
In an orthonormal reference, a periodic elec- Px =
0 0
tric eld, which we have expressed in a ref-  
erence formed by the diopter, and the plane 0 0
Py =
of incidence can be reduced to the following 0 1
60 Chapter 6 Physical Colors, Chemical Colors Basics of Solid State Optics

Rotator. Some structures, commonly found


in Coleoptera elytrons, make polarization ro-
tate. Jones matrix then takes the form of the Optical Index
traditional rotation matrix R. As has just been seen with the few previous spe-
  cic cases, the optical index is the fundamental
cos sin
R= parameter of electromagnetic waves. We will
sin cos
now put waves aside for a moment in order to
focus on matter and its optical properties char-
acterized by its index.
One should rst gure the various phenom-
ena that can occur in a material when it is sub-
ject to the action of an electromagnetic wave.
We will limit our eld of study to material types
that can be found in biological structures. We
Dephasor. Finally, some anisotropic struc- can thus exclude metals, semiconductors, and
tures can produce a phase change  between ionic crystals and focus on the electromagnetic
the eld components.  is related to the differ- eld action. Its interaction with matter occurs
ence in the optical paths  of the two compo- through the electric chargeselectronsand
nents by: produces different phenomena.

2 r Firstly, in every atom, the periodic electric


= . eld of the wave makes electrons move away

from their initial position and deforms their
  orbitals, thus generating oscillating dipoles.
ei/2 0 r Secondly, in a molecule, the eld affects the
=
0 e+i/2 links between atoms, shortening or lengthen-
ing them, changing their orientation . . .

Figure 6.5. After crossing an


anisotropic material, the vertical
component has a quarter of wave-
length optical path difference with
the horizontal component. The E
extremityor its projectionhere
follows an ellipse.
LightMatter Interaction, Polarization, and Optical Index 61

Figure 6.6. Action of a peri-


odic electric eld E() oscillat-
ing on atoms or molecules in
the different spectral ranges.

r The eld can nally tend to orienton inary part characterizes the absorption of the
averagemolecules containing a permanent material.

electric dipole in a given direction. = 1 + i2 ,
(6-5)
In any case, there is a displacement of n = n ik.
charges. In electromagnetism, this displace- We wont elaborate on the calculation of the
ment is called D. As long as the applied eld dielectric function, which doesnt concern us
is not too stronglike with sunlight and tradi- directly. A simple method consists in consid-
tional electric lampsthe displacement is pro- ering the charges in motion as small masses
portional to the electric eld: carrying electric charges linked to the nucleus
or to neighboring charges by springs: This is
D = E. (6-3)
the mechanistic model, basic but allowing us
The proportionality constant is a character- to understand most of phenomena. For cer-
istic of the material, or rather of its response to tain specic frequencies, 0 , this oscillator res-
the action of light. Since it depends on the wave onates and the absorption is maximal. The
frequency, it is called dielectric function of model leads to the following expression of the
the material. By denition, the optical index is dielectric function:
the square root of : Ne2 /m
 () = n2 () = 0 +  2  ,
n() = () (6-4) 0 2 i
() = 1 () + i2 () and n() = n()ik().
This mere denition doesnt tell us anything
about the aspect, the value, or the frequency (6-6)
dependence of this index . . . Let us now con- N is the number of oscillators, i.e., the num-
sider the latter, with the least equations pos- ber of molecules or polarized atoms per unit
sible. It is important to note rst that if the volume, and is a damping term, both param-
displacement D (i.e., the effect), is indeed pro- eters characterizing each kind of atoms. As il-
portional to the applied eld E (i.e., the action), lustrated by Figure 6.7, the 2 imaginary part
they are not systematically in phase; a eld can variation, directly linked to the optical absorp-
vary very quickly, while a charge can be slowed tion, looks like resonance peaks centered on
down in its motion. the various 0 frequencies. As concerns the
This phase difference is expressed by making 1 real part, it tends towards the static dielec-
become a complex quantity with a real part, tric constant (0) of the material at low fre-
1 , and an imaginary one, 2 , which directly ac- quency (  0 ), whereas at high frequency,
counts for the energy provided by the wave in it tends towards that of vacuum by decreasing.
order to move charges, i.e., the energy absorbed This gradual decrease of the real part of the di-
by the material. Like the dielectric function, the electric function of insulators is important and
index n also becomes complex and its imag- worth mentioning. One can show that the real
62 Chapter 6 Physical Colors, Chemical Colors Basics of Solid State Optics

and imaginary parts of the dielectric function dielectric function, but that, on the contrary,
are not independent, and that they are related every integral has a positive value at low fre-
to each other by integral relations: Kramers- quency (  n ) and their different contribu-
Kronig relations. tions n are additive and approximately equal
to:
 2 ( ) 

1 () 0 = 2 2 d ,
2 
2 2 ()
0
(6-7) 1 n d . (6-9)
 [1 ( )0 ] 

2 () = 2 2 d 
. picn
2
0
In other words, each absorption peak con-
I decided to mention these relations here be- tributes to 1 at low frequency, and so does
cause they have an important physical mean- the index real part n, which is always larger
ing. First, with one componentgenerally 2 , in the red side of an absorption peak (low fre-
which can be more easily determined through quency) than on the blue side (high frequency).
experimentone can calculate the other one. This is nothing than the mathematic expres-
Second, these relations express the index dis- sion of a well known phenomenon: The larger
symmetry on both sides of a resonance, and the mass, the more difcult to make it oscil-
therefore selective reection, i.e., surface color, late at high frequency. In the case concern-
that was so dear to Michelson. ing us, when frequency is gradually increased,
As has been seen, the 2 imaginary part of the molecules, atoms, and electrons stop vibrating
dielectric function looks like peakspossibly successively. At each step, the dielectric func-
severalcentered on a n frequency, and be- tion decreases to nally reach the value of the
tween these peaks, 2 is almost equal to zero dielectric function of a material in which noth-
(6.7). The integral resulting in 1 as a function ing vibrates: vacuum (0 ).
of 2 , which is a sum on all frequencies can Let us nally note that the index varies sig-
thus be expressed as a series of integrals over nicantly only when it is close to resonance
the only ranges where 2 is not equal to zero, frequencies. Outside of these high absorption
i.e., on the various absorption peaks: areas, within transparency windows, the index
 2   2 ( ) real part is not strictly constant and slightly
1 () = 0 + 2 2
d . varies with wavelength. Rather than using ex-
n
picn pression (6-6)which still remains valida
(6-8) better account of the variation is given by an
One thus notes that at high frequency ( empirical expression. This polynomial expres-
n ), all of these integrals are equal to zero sion as a function of , or , is easier to use. Its
and 1 tends towards 0 , demonstrating that coefcients are determined by tting the exper-
the various absorptions dont take part in the imental variations of the index of each material.

Figure 6.7. Evolution of the real part


of the dielectric function of a material
presenting two absorption peaks. Each
peak brings a nite contribution on its
red side, towards low frequencies. 1
is always larger on the red side of an
absorption than on its blue side.
LightMatter Interaction, Polarization, and Optical Index 63

inclusions are generally very small and with


low concentration, so that the medium is highly
dissymmetric on the one hand. And on the
other hand, a matrix can be dened without
ambiguity, the m dielectric function embed-
ding isolated and noninteracting inclusions of
i dielectric function. We will always be far
from the percolation regime (which we wont
Figure 6.8. Denition of an effective medium by the
mean eld theory. The eld in the effective homoge- mention) where inclusions, more numerous,
neous medium is the volume average of the elds in tend to join together and form aggregates. We
each component. will focus on the simpler case where inclusions
are much smaller than the wavelength, so that
the electric eld isas a rst approximation
Effective Index
uniform on the whole inclusion (in which case
In what has previously been seen, the materi- the regime is referred as quasi static). The
als considered were implicitly homogeneous, method traditionally consists in expressing the
i.e., containing no heterogeneity nor inclusion Ee effective eld and the De effective electric
larger than the characteristic size of molecules. displacement in the composite as the volume
This is very rare concerning biological materi- average of the elds and displacements taking
als and we will later encounter media consist- place in the inclusions with volume concentra-
ing of a continuous matrix containing a certain tion p on the one hand, and in the matrix on
number of inclusions with various indices. In the other hand.
this way, we have to determine the index of 
Ee = pi Ei + (1 p)Em ,
such a disordered medium, after making sure (6-10)
De = pi Di + (1 p)Dm ,
that such an index has a physical meaning.
Any real medium presents heterogeneities, The second equation can be expressed as fol-
the typical sizes of which can vary from atomic lows:
scale (alloys, defects . . . ) to macroscopic scale e Ee = pi Ei + (1 p)m Em . (6-11)
(composites, concretes . . . ). This classication
is not as simple as it seems; any such inhomo- Traditional electrostatic calculation allows to
geneous medium that is scattering in the visible link the Ei and Em elds and thus to dene e in
or ultraviolet is not scattering anymore at lower relation to p, i and m . For spherical inclusions,
frequency, thus presenting a homogeneous be- one obtains:
   
havior. Therefore, the domains are determined i 1 + 2p + m 1 p
according to the wavelength but not in the ab- e = m    , (6-12)
i 1 p + m 2 + p
solute.
A homogeneous medium is indeed charac- This expression was established by Maxwell
terized by a rectilinear propagation of light. Garnett in 1904 and was named after him.
In an inhomogeneous medium, when light This average medium presents interest-
encounters inclusions, with sizes comparable ing properties, especially when inclusions are
to the wavelength, a scattering regime takes metallic. In this case indeed, the real part of the
place. We will further describe this regime, denominator of expression (6-12) can vanish
which is widespread among butteries. If within a specic range of frequencies, generally
on the contrary, inclusions are smaller than in the visible, leading to a high absorption
the wavelength, there is almost no scattering generating colors. This is how stained-glass
and the medium behaves like a homogeneous windows colors are obtained. Yet, this effect
medium. One can thus try to deduce the e av- doesnt concern butteries, so we will leave
erage or effective dielectric function of these it aside. Among the latter on the contrary, the
homogenizable composite materials. In the materials making the compound are often of
materials that we will encounter in butteries, the same typedielectricsand present close
64 Chapter 6 Physical Colors, Chemical Colors Basics of Solid State Optics

index values. One can thus show that the vol- quently the dipole value depend to a great ex-
ume average of the dielectric functions rep- tent on the inclusion geometry. One thus intro-
resents a good approximation of expression duces a shape factor, called depolarizing factor
(6-12). This is the linear theory of mixing, which A, which modulates the intensity of the dipole
we will be able to apply to butteries in most generated by the eld. It is generally difcult to
cases: calculate this coefcient. It leads to integral ex-
pressions with no analytic solutions, which can
e = pi + (1 p)m . (6-13)
be determined only numerically thanks to aba-
Maxwell Garnetts research focused on the cus. In the more common case of spheroids and
optical properties of colloidal suspensions of other simple geometries, these elliptic integrals
metals and on the colors of stained-glass win- are expressed by using basic functions de-
dows obtained from metal inclusions in glass. pending on the inclusion eccentricity. We are
It is to this phenomenon, among others, we only concerned with their graphic represen-
owe the extraordinary blues of Medieval Age tations and more especially the extreme cases
stained-glass that illuminate our cathedrals. In of long innitely inclusions, like chitin sticks,
this specic case, the quantity of metal inserted or extremely at inclusions, like stria lamellae.
is extremely small and metal atoms, lost in a These shapes are the closest to that observed in
dielectric ocean, join together and form small our biological structures. The coefcient varies
aggregatesseveral tens of nanometers big from the value A = 1 for a at inclusion to A =
that are somewhat spherical at this stage. In 0 for a very elongated inclusion when the eld
order to use this approach in the framework is oriented in the direction of the revolution
of the biological structures concerning us, one axis. In the perpendicular direction, it is equal
must consider nonspherical inclusions. This to half of (1-A), i.e., 0 in the rst case and 1/2 in
development of the model was started more the second. Those are the only values needed
than half a century later. When an inclusion is to model optical structures of insects.
subject to an electric eld, a displacement of A mean eld calculation according to
charges systematically occurs. This results in a Maxwell Garnett results in the following
charge accumulation on the periphery of the in- tensor expression of the dielectric function,
clusion, which thus behaves, as a rst approx- established by Cohen et al. in 1973:
imation, like an oscillating dipole if the eld is    
i A + p(1 A + m 1 p (1 A)
periodic, which is the case in optics. One can e = m     ,
intuitively understand that the distribution of i A 1 p + m 1 A(1 p)
charges on the inclusion periphery and conse- (6-14)

Figure 6.9. Depolarization coefcient of


a spherod in relation to the a/c axes
ratio. The inductor eld is parallel to
the revolution axis a. Non spherod
geometriesplane and cylinderhave
the same depolarization coefcients as
that of very at or elongated spherods,
respectively.
Structural Origin of Colors; What Kind of Logic for Structures? 65

Figure 6.10. An electron moving along


a 1-dimension crystal is subject to
a Coulomb periodic potential that is
schematically represented as a crenal po-
tential.

This tensor expression reduces to the where is the wave function of the particle and
Maxwell Garnetts scalar expression when A = E its energy. We wont try to solve this equation,
1/3, which is depolarizing factor of a sphere. It which is actually not so easy in the present case.
is important to note that for a given structure, e We will rather keep in mind that its solutions
wouldnt have the same value for two different are discontinuous; propagation cannot occur
orientations of the electric eld. This is partic- with any energy. Allowed energies are grouped
ularly true in Morphos, when the structure is in bands separated by energy areasforbidden
illuminated with Transverse Electric or Trans- areasthat the particle cannot experience. The
verse Magnetic polarized light. This anisotropy potential periodicity causes a partial quanti-
of the material index results in both colors and cation of energies.
polarization effects observed in Morphos. Let us now come back to optics, and more
precisely to buttery structures. The particle is
now a photon and its associated wave is the
Structural Origin of Colors; What electromagnetic wave. What happens when the
Kind of Logic for Structures? particle tries to propagate within this medium
presenting a periodic alternation of index, the
The Ordered Phase; Photonic Crystals equivalent of the electron electric potential?
Before Time Strictly the same thing happens! The equation
to solve is now Helmholtzs:
In the traditional sense of the term, the crys-
tal phase characterizes a periodic organization n2 2
2 = 2 , (6-16)
of atoms both at short and long distances. In c
any place on the crystal, one can observe that which is expressed in the exact same way
atoms are distributed in the same way. This as Schrodingers equation (the frequency
strict periodicity has surprising consequences replacing the E energy) and leads to the same
on particles trying to move within the crystal. type of solutions: successive bands of per-
In order to solve such a problem, quantum me- mitted frequencies alternating with bands of
chanic starts by strictly modeling the shape of forbidden frequencies. The index periodicity
the potential subjecting the particlegenerally causes a partial quantication of the frequency
an electronwhen the latter comes near atoms. range.
This approach substantially simplies calcu- The term photonic crystal proceeds from
lations, without losing any physical meaning, this analogy between electrical phenomena in
and gets us closer to the optical phenomenon a crystal and optic phenomenaalso called
that are interested in here. photonicin a structured medium.
One then has to solve Schrodingers famous It is in this context that we will describe the
equation, which in this case is expressed as fol- structures of the wing and the associated phe-
lows: nomena. From the point of view of geometry
indeed, one can classify photonic crystals ac-
2m cording to the number of dimensions in which
2 = - 2 (E V), (6-15) periodicity develops.
h
66 Chapter 6 Physical Colors, Chemical Colors Basics of Solid State Optics

Figure 6.11. A photon with energy E = -h prop-


agates into a stratied
medium. It successively

goes through high 1 and low 2 index
materials.

One-dimensional crystals present a periodic


index following one direction, and is uniform
according to the two others. It is the case, for
instance, for overlapping thin layers of alter-
nate high and low indices. One knows that
such structures produce interferences, and are
to a large extentas they have been for a long
timeused as dielectric mirrors, lters, and
so on. In the same way, 2-dimensional and

Figure 6.13. Two amorphous structures; (a) defaults


introduced in a periodic structurethe distances
separating elements are a multiple of the crys-
tal period. (b) The distances between elements are
random.
3-dimensional crystals present periodicities in
respectively two and three dimensions, corre-
sponding in this last case the traditional geom-
etry of mineral crystals.
Figure 6.12. Three examples of a 1-dimension (a) These gratings are the origin of phenomena
2-dimension (b) and 3-dimension (c) crystalline that have historically been classied as diffrac-
structure. tive but that are actually a generalization of
Structural Origin of Colors; What Kind of Logic for Structures? 67

Figure 6.14. In both gures, the phase


difference between the three waves is a
multiple of . (a) Their shift is inferior

to the coherence length d, the overlaps
are large and waves can interfere. (b) The
shifts are larger than d, the waves do not

p
la
overlap and cannot interfere.

er
Ov
interference phenomena encountered in a thin ing the two examples presented in Figure 6.12,
layer. disorder can proceed from two different ways.
After studying ordered structures, or crys- The method that may be the more intuitive con-
talline phases, let us now see what happens sists in arranging elements at random (Figure
when this beautiful arrangement deteriorates, 6.13 (b)). One can easily understand that there
i.e., when it gets in disorder and when the phase is no phase relation between the two waves
becomes amorphous. diffracted by each of the elements. In fact, such
a structure doesnt lead to any diffracting phe-
Amorphous Phase nomenon.
One can also, as shown on Figure 6.13 (a),
All the phenomena that have been mentioned
randomly remove elements from an ordered
(interferences, diffraction) occur because there
structure. There is no order over a short or long
is a relation between the phases of the two
distance anymore. It is indeed an amorphous
waves diffracted by two neighboring objects,
phase, although all the distances between
and because the periodicity of the two ob-
elements remain a multiple of the initial crystal
jects enables this relation to spread to all the
period. In a given direction, diffracted waves
waves, thus leading to a global effect that can
can thus very well be in phase, actually out of
be observed. Were the periodicity lost, the ef-
phase of n wavelengths the one to the others,
fect would vanish. Phase relations would still
with n varying randomly. Consequently, noth-
occur here and there, but they would not pro-
ing prevents interference phenomena from tak-
duce any global phenomenon. This is amor-
ing place. Yet, this hardly occurs with natural
phous phase. Light is scattered. As the beau-
light. The reason is quite simple: waves are not
tiful arrangement has been lost, only the sizes
innite but truncated. They have a beginning
of the scattering objects and the compactness
and an end. One thus speaks of trains of waves,
of the whole play a part. This is this type of
and the nite length of these trains is called .
structure that produces most of the blue and
If the distance between two objects is superior
white colors found among butteries in our re-
to this length, then the two resulting waves
gions. The very ne structures of Argus pref-
even if they were in phasedont superpose
erentially scatter blue, whereas the structures
and therefore cannot interfere. We will call co-
of bigger ovoid grains of Pieridae scatter all
herence length, this distance beyond which two
wavelengths, hence their whitish color.
waves in phase dont interfere anymore. Let us
only remember that a periodic structure is not
An Important Parameter: Coherence Length
necessarily sufcient to produce constructive
Disorder can be seen as quite an intuitive no- optical phenomena and that the latter can only
tion, but it is actually very complex and we proceed from ranges that are smaller than the
wont tackle it. Let us only note that consider- coherence length.
7
1-Dimensional Structures:
Interferences

Theory Recalls the indices of the two media separated by the


interface.
Interference in Thin Layers The equations of continuity on the eld
phases indicate their propagation directions.
Interferences in thin layers are common but The reection angle r is equal to the inci-
produce colors that are not very bright (soap dent angle i , and the angle of the transmit-
bubbles, oil or gas sheets . . . ). The intensity can ted waveor refraction anglet is linked to
be increased by stacking many layers. Multi- the incident angle through the relation involv-
layer systems are indeed commonly used in the ing the medium indices ni sini = nsint. . These
industry but also in nature. two laws are known as the Snell-Descartes re-
An optically homogeneous medium is char- lations.
acterized by its complex index n = n ik (k As for the equations of continuity on ampli-
being equal to zero when materials are not tudes, they lead us to Fresnels formulas, which
absorbing), the physical sense of which has link the reection and transmission coefcients
just been mentioned. We will follow the evo- of the incident wave (s and p components) to
lution of a plane wave from the entry to the the medium indices:
exit interface of a thin layer. We will thus suc-
cessively examine a simple interface between ni cos t n cos i

rp = ,
two different media, the front medium being ni cos t + n cos i
and
usually air, with a real index ni = 1. We will
n cos i n cos t
rs = i ,
then study the second interface, the inferior ni cos i + n cos t
face of the thin layer. This will be the end of

our physical study, the multilayer represent- 2ni cos i
ing a mere repetition of the same elementary
tp = ,
ni cos t + n cos i
problem. (7-1)

2ni cos i
ts =
ni cos i + n cos t
The Plan Diopter
Let us consider an electromagnetic wave falling Under normal incidence (i = r = t = 0),
on an interface between two semi-innite me- every wave is parallel to the layer plane, and
dia. Part of this wave crosses the diopter the preceding relations become:

this is the transmitted wave. Another part, ni n
the reected wave, goes back to the incident
r = n + n ,
i
medium. On both sides of the interface, the . (7-2)

2ni
electric and magnetic elds are not indepen- t =
ni + n
dent, as they come from the same wave. Bound-
ary conditions impose very specic propaga- One more commonly uses the reection and
tion directions and amplitudes, depending on transmission factors, ratios of the reected and

68
Theory Recalls 69

Ep r

Es
i r i
ni
n I K
e n
r

t J
t

Figure 7.1. A diopter and its normal form an orthog- Figure 7.2. The optical path of the rst rays reected
onal coordinate system in which any wave can be and transmitted by a thin layer. The intensities of
split into two waves vibrating respectively in the the rst two reected rays are almost similar. The
plane of incidence (Ep ) and in a plane parallel to the intensities of all the transmitted rays, excepted the
layer (Es ). rst one, are negligible.

transmitted intensities with the incident inten- direction than in its amplitude falling on the
sity, which can be directly obtained from spec- layer inferior side.
trometric measurements: The Thin Layer
Re {n} 2 The wave reected on the rst interface is ob-
R = |r|2 and T = |t| (7-3) viously not affected by the presence of the sec-
ni
ond one, but the transmitted wave now plays
In the case of butteries in a weakly or non- the role of an incident wave with a different an-
absorbing medium like air, which we will often gle t . It here splits, following the same laws at
encounter, the reection and transmission fac- the superior interface into a transmitted wave
tors become: of high amplitude and a reected wave of low
   
 n 1 2  2 2 amplitude. The latter, back on the upper in-
R=   and T = n    . (7-4)
n + 1 n + 1
terface, splits again into a transmitted wave
and a reected wave. Although this play of
Before going further, let us note that indices come and go between the two faces of the layer
of the most widespread nonabsorbing natu- goes on innitely, producing an innite num-
ral/biological materials (for example, water ber of transmitted rays, it is not necessary to
and chitin) are not very highapproximately analyze it further. By following point by point
4% in the mentioned casesand that most of the trajectory of the rst transmitted wave and
the incident wave is transmitted. It is therefore the evolution of its amplitude, one can easily
a wave that is more affected in its propagation observe that the rst two reected rays have

Figure 7.3. The evolution of con-


structive interferences according
to the layer thickness. (a) For the
given thickness and incidence,
red waves are in phase after re-
ection on the two interfaces of
the layer: interferences are con-
structive. (b) If the thickness is de-
creased, this behavior is followed
by the blue waves.
70 Chapter 7 1-Dimensional Structures: Interferences

approximately the same low intensity, whereas All basic physic phenomena are contained
that of the following reected rays are insignif- in the treatment of the unique layer and to
icant. In the same way, only the intensity of the mathematically analyze the multilayer would
rst transmitted ray is important; interference not provide us with much more information. It
phenomena proceeding from the reection on only deprives us of the pleasure to present my
thin layers can be considered as phenomena teachers, F. Abeles, delicate matrix treatment,
of two waves with almost similar intensities, in which every layer is represented by a matrix
and thus of maximum contrast. These two rays, linking the elds at the entry and at the exit,
generated by the same incident ray, are coher- the effect of the multilayer being obtained by
ent and interfere. In a given direction, these in- multiplying matrices.
terferences are constructive only for a given
wavelength, which indeed gives the wanted Summary
iridescent effect in the visible. The main results, as derived from the study
The period difference between the two rst of the unique layer or multilayer are: (a) there
reected rays is: is no absorptionthe multilayer interferen-
tial mirror is not a colored lter in the lit-
= 2ne cos t + . (7-5) eral meaning of the wordand nonreected
2
wavelengths are transmitted. The color of the
The latter are thus out of phasedestructive
transmitted light is thus the complementary
interferencesif d is equal to an odd number
of that of the reected light. (b) As shown
of half wavelengths:
by the equation (7-8), the wavelength re-
ected by a multilayer mirror varies linearly
(2k + 1) = 2ne cos t + (7-6)
2 2
For a given conguration, there is a minimal
reection for each wavelength verifying

kmin = 2ne cos t , (7-7)

where k is an integer, and the reection maxi-


mal for:

kmax = 4ne cos t , (7-8)


i.e., under normal incidence, for optical thick-
nesses equal to a quarter of wavelength.
The Multilayer
As has been seen, the reection factor gen-
erally remains weak. Refraction indices that
are widespread in nature span from 1,34 for
cytoplasm to 1,83 for guanine crystals. Un-
der normal incidence and for an optical thick-
ness of /4, such a thickness doesnt reect
much more than 8 to 9% of the incident en- Figure 7.4. The optical path in each layer of the stack
ergy for this wavelength. One can nevertheless must be equal to the quarter of the incident wave-
increase quite signicantly this reection fac- length. The waves reected by each interface are
phase synchronous. This occurs when ni el = nh eh =
tor by multiplying the number of layers, i.e.,
/4. The more important the index difference nl nh
by piling a certain number of layers of high between the two materials, the larger the amplitude
and low indices with optical thicknesses close of the reected waves. Ray deviations caused by
to /4. crossing each diopter are not gured here.
Theory Recalls 71

structure can be found in nature. The sec-


ond phase can be gaseous, generally air, like
among butteries and birds, or solid among
Coleoptera, sh and shell, or even but rarely
liquid in cephalopods. We will also mention
the possible changes in phase, a liquid replac-
ing a gas in the structure, leading to a variation
in optical thickness, and consequently of the
reected color.
Compared to pigmentary colors, and except-
ing colors obtained through scattering, phys-
ical colors are bright and rather pure. From
the sole point of view of intensity, the mes-
Figure 7.5. Variations of interferential colors with
the angle of incident. For given thicknesses and sages delivered, whether sematic or apose-
indices, the wavelength of the reected wave de- matic, are optimized. This requires that there
creases when the incident angle increases; color is no dead angle, if I may say so. However, a
tends towards blue. thin layer is a mirror, even if it selects wave-
lengths, as it reects light in a specic direction.
Interferential devices must therefore be asso-
with the cosine of the angle of incidence, ciated with a system of spatial dispersion of
and thus decreasestending towards blue light.
when the latter increases. (c) Lastly, the more As we will see, such systems are very often
constant and even the thickness of the vari- present, whether on wingsColeoptera and
ous layers, the purer the reected color. One their convex elytronsor on scales, or even at
can almost obtain white mirrors by gradu- a smaller scale on mesoscopic structures in the
ally modifying the thickness of the layers in a scales. In any case, the diopter is deformed.
multilayer. Light comes under a variable incidence and
the dispersion is generally associated with po-
Various Types of Multilayer larization effects. This very principle will help
In interferential phenomena, materials are us classify and present interferential structures.
involved only through their indices. They The deformation of the diopter results in con-
can thus occur, whatever the phase of the cave and convex structures, and more rarely
componentsliquid, solid, or gaseouspro- to structures alternatively concave and convex.
vided the required conditions are respected. We will begin our study with convex struc-
Almost every kind of combination involving tures, from the biggest ones to the smallest
at least a gaseous phase and supporting the ones.

N = 10 N=5 nh = 1.56
0.8 nl = 1.00

0.6 N=2
0.4

Figure 7.6. Variations in the reected in- 0.2 N=1


tensities with the number of layers, for a
given system. Indices are approximately
equal to that of airnl = 1and chitin 400 500 600 700
nh 1,56.
72 Chapter 7 1-Dimensional Structures: Interferences

(a) (b)

(c)

Figure 7.7. Some examples of interferential colors. (a) a solid-solid multilayeran Aliotis Iris nacre. (b) a
liquid-liquid multilayer; a thin layer of petrol over a water lm. (c) a solid-gas structure quite similar to
that of Butterieshummingbird feathers.

Type Butteries with Convex after them: Morpho type and Urania type scales.
Structures Indeed, they are big butteries presenting areas
of rather homogeneous colors and are well-
Chrysiridia madagascariensis, Urania leilus, suited to routine spectroscopic measurements.
and Prepona We will comply with this convenient tradi-
tion. Yet, facility here is only apparent since
Interferences and diffraction are the two physi- at the microscopic scale, the two families do
cal phenomena leading to the most spectacular present two specic types of scales. Yet con-
iridescent effects. They occur among miscella- cerning Uraniidae, their scales are quite de-
neous buttery species throughout the world, formed, which makes it difcult to get a gen-
yet hardly with the same brightness as the large eral view, and as concerns Morphidae, the effects
neotropical family of Morphidae and the more produced at the smallest scale are then modu-
smaller one of Uraniidae represented by the lated.
Madagascan species Chrysiridia madagascarien- We will start our study of iridescence
sis and South-African Urania leilus. as produced by convex structures of Urani-
For practical reasons, the study of colored idae (mostly Chrysiridia madagascariensis) and
effects has always been carried on these two Charaxinae (Prepona), which provide a better ba-
families. Certain scales have even been named sis for the interpretation at a small scale. We
Type Butteries with Convex Structures 73

Figure 7.8. The South-African UraniaUrania leilus.

will discover those butteries by plunging into


the innitely small, according to six different
orders of magnitude, and zooming out, estab-
lishing each time the principles of the optical
effects are produced or suppressed.
As we have already seen, iridescence can be
created either by variations in the incident an-
gles of the light ray or by changes in the view-
point of the observer. That is why it may be use- Figure 7.9. Iridescence in Chrisiridia madagas-
cariensisunder normal incidence (0,0) above and
ful to specify the meaning of our notations at in low-angled incidence (90,0) below. In the anterior
the beginning of this chapter. Except for contra- wings medial area, the predominant color goes from
dictory cases clearly outlined, the observation 565 nm (yellow-green) to 505 nm (green) and reaches
of the whole buttery as well as scales under 495 nm (blue-green) in places. What is remarkable is
the optical microscope is made perpendicular that effects are similar on the ventral side
to the plane of the wings or the membrane. It is
the light source that is mobile and its position green to yellow for the anterior wings and
is traditionally established by its azimut and red to blue for the posterior ones. Basal and
its declination. marginal areas of the dorsal side are deep black
(Uraniidae warm themselves in the dorsal posi-
tion), and the posterior is entirely fringed with
Uraniidae and Charaxinae
long white scales. All those colors vary with
The Uraniidae Chrysiridia madagascariensis is a incidence, and once dipped into trichloretylen
Madagascan crepuscular buttery that lends (index: 1,478), the buttery shows a regular
to breeding. Its bright wings are unfortunately brown color, which demonstrates the structural
extensively used in handicrafts. The colors of origin of colors and the systemic presence of
their medial areas on both faces cover almost melanin in the scales of both faces. The disap-
the whole range of the visible spectrum from pearance of black suggests that structures also
74 Chapter 7 1-Dimensional Structures: Interferences

Figure 7.12. Typical arrangement and sizes of


Chrysiridia scales. Structural scales emerge almost
horizontally from cupules and undergo a double
curvature before covering the roots of the preced-
Figure 7.10. Diffuse reection factor of the green ar- ing row of scales.
eas of Chrysiridia madagascariensis anterior wings
Cu1. The role of the wing membranes and of the
scales themselves, which is signicant in infrared,
has been underlined in order to reveal that of the depends on the membrane. These proper-
scales structures. ties are quasi-constantly found among all
species.
Scales are consequently the driving point of
play a part, at least as important as that of pig- the reecting structure in the visible eld, from
ments in the emergence of blacks, hence in vis- which colors proceed. The shift of the peaks in
ible absorption. the visible spectrum observed under the optical
Reection spectroscopic measurements car- microscope when varying the incidence (Fig-
ried on the green areas of the anterior wings ure 7.11) demonstrates that colors are interfer-
show fringes that are characteristic of in- ential and thin layer-type. The high value of the
terferential effects in thin layers. The same reected intensity (more than 30% at 560 nm, at
measurements conducted on the sole alary the limit between green and yellow) suggests
membrane, from which the scales of the two the presence of a multilayer structure given
faces were removed, allow us to point out the the concerned materials. Optical microscopy
roles of the various components in the general shows regularly implanted scales arranged in
reection of the wing. We can thus conrm two different layers. The rst layer, consist-
that scales are in charge of the visible struc- ing of the bottom pigmentary scales folded
tures, whereas most of the infrared reection longitudinally and slightly convex, form an

(a) (b)

Figure 7.11. The two scales types of Urania leilusin arrangement (right) and isolated (left). Ground scales
form a black screen under cover structural scales that present an interferential color variation from the crest
to the root.
Type Butteries with Convex Structures 75

(a) (b)

Figure 7.13. SEM views of the dorsal side of Chrysiridia madagascariensis, (a), and of Urania leilus, (b). They
show the general arrangement of structural scales and their signicant convexity. One can distinguish
pigmentary ground scales under the indented apex.

absorbing screen on which the second layer of two types of scales present not very distinct
structural scales is placed. The latter, strongly longitudinal striae, with a larger spacing on
convex, slightly pigmented, and their sides ad- the structuralabout 4 m apartthan on the
joining, form long rolls resembling waves, on pigmentary ones (1 to 2 m). There are al-
the crests of which the light is reected, thus most no counter-striae on the structural scales,
drawing long colored bands transversal to the while pigmentary scales present some that are
wing (Figure 7.18.). An enlargement with scan- very distinct and regular with spaces from 230
ning electron microscopy enables us to under- to 300 nm. It is this quite condensed alveo-
line the general shape of the scales as well as lar structure, which is responsible along with
their supercial structures (Figure 7.12). The melanin for the strong visible absorption of the

Figure 7.14. A Chrysiridia


madagascariensis structural
scale above and a pigmentary
ground scale below. The
alveolar structure formed
by striae and counter-striae
shows up very clearly on the
latter. It signicantly increases
visible absorption. Over the
structural scale striae, the
spacing of which is much
bigger, one can note the
overlap of very short lamellae
inclined in a large extent on
the membrane plane.
76 Chapter 7 1-Dimensional Structures: Interferences

Figure 7.15. Transmission Elec-


tron Microscopy image of the
cross section of a Chrysiridia mada-
gascariensis structural scale. The
outgrowth in the center is a
stria, which apparently presents
a pleated structure similar to that
of Morphos, but too small to pro-
duce any effect in the visible.

ground scales, that we will further describe in croscopy (TEM) reveals the superior mem-
the study of the thermodynamic properties of brane consisting of 13 to 15 overlapping al-
butteries. ternate layers of air and chitin, 90 nm thick
Let us now observe the striae of the struc- on average, and supported by microtrabeculae.
tural scales. They show a traditional structure TEM also reveals the external layer presents
composed of short lamellaes that dont over- many pores. The penetration speed of the
lap, the length of which is almost equal to various index liquids in the scales suggests
their spacing. Given their structure and spac- that the deeper layers should also include
ing, they cannot be the source of the strong pores.
reection that one can observe. One must get Between two striae (that is, on distances
into the scales in order to discover the se- about 4 m), the structure is rather at and is
cret of the Urania. Transmission electron mi- easily modeled by a symmetric multilayer of

Figure 7.16. Cross section of a Chrysiridia madagascariensis structural scale cut almost through the middle.
The spacings between two striae, slightly concave, can be likened to planes tilted of about 10 in relation to
each others.
Colored Effects 77

Figure 7.17. Variations of colors and of polarization


(a) and schematic representation of an Archeoprepona
structural scale. (b) A ray falling under near-normal
incidence onto the center of the scale generates a
non polarized color. The same ray falling closer to
the apex or the root produces a bluer color (higher
incidence) and more polarized (incidence close to
the Brewster angle.)

chitin and air of about the same thickness and


without any support. A t of the visible reec- Figure 7.18. Among all butteries with convex
tion results in index values around 1,6 for chitin scales, only a part of the scale reects light in a given
and slightly larger than 1 for the air layer, due direction, which results in color stripes. Prepona an-
terior wings.
to the presence of microtrabeculae in it. The
scales being a little convex transversally, one spatial dispersion of light. They are conse-
can also assume that in this direction, it looks quently always associated with a dispersive
like a mirror consisting of 4-m wide facets, system. When the convexity of scales is respon-
each inclined by an angle of about ten degrees sible for dispersion, as we have just seen with
with respect to the next, slightly more in the the Urania, incidence varies from the center to
center and less on the periphery. This tends the edge of scales, causing the reected light
to broaden the reection peak, which yet re- to disperse substantially with two coexisting
mains centered on = 560 nm under normal effects. On the one hand, the incident angle in-
incidencea value corresponding exactly to creases from the center to the base of the scales
what was theoretically assumed (equation 7-8) apex leading to the reected wavelength shift-
resulting in an average thickness of 87.5 nm ing towards the shorter wavelengths, hence the
and to decreases the spectral purity. The same blues. On the other hand, certain areas are illu-
effect is produced in the longitudinal direc- minated under the Brewster incidence, which
tion of course, yet substantially augmented, the makes the corresponding color highly polar-
scales almost forming a semicircle so that only ized. Let us note that (Figure 7.17) the various
a small area in the crestabout 50 microns colors are not reected in the same direction
can produce interferences in the visible. Spec- and so dont mix. This is an iridescent effect,
tral purity is here again reduced. which is associated with a polarization effect.

Dispersion and Polarization Colored Effects


Multilayer systems can generate bright and As we have already mentioned, structural
rather pure colors, but cannot produce the scales are usually well organized. This is
78 Chapter 7 1-Dimensional Structures: Interferences

strictly speaking. Iridescent concave structures


are found at a smaller scale, especiallyas Lep-
idoptera are concernedin the big family of
Papilionidae and among Cincidelae that belong
to Coleoptera.
Hemispheric concave diopters belong, along
with perpendicular plane mirrors, to optical
structures that reect light under its direction
of incidence, whatever the incidence. It is thus
not surprising that these structures produce re-
markable iridescent effects, since there is no
shading on scales anymore, like in the previous
case. Papilionidae are indeed among the bright-
est Lepidoptera.

Papilionidae: Solid/Gas Structures


The general arrangement of scales of Pa-
pilionidae of the genus papilio is similar to
those of Urania and Prepona. In colored areas,
mostly green or blue, cover scales are struc-
tural and the ground scales pigmentary. Nu-
merous species of the genus are characterized
by structural scales that are scattered over the
entire wings, which are very dark in addition.
This produces a very beautiful effect of span-
Figure 7.19. Scattered structural scales on the Papilio gles that are particularly perceivable under a
paris black wings. rather high incidence.
Structural scales all present almost the same
shape; slightly rectangular and convex with
particularly noticeable in convex structures un- rounded angles. Striae are spaced (around
der direct light. The crests of scales illuminated 5 m) and little structured. Inter-striae spac-
under near-normal incidence create long and ing is regularly undulated with a period
very bright bands contrasting with the con- and amplitude varying according to species.
stantly blue background formed by the base The striae and undulations in the membrane
and apex of the scales illuminated under low- thus isolate basins, more or less long and
angled incidence. deep.
The green color of Uranidae in particular is Cross-sections observed through transmis-
due to a mix, not pointillist, but of bands, green- sion electron microscopy show a stacking of
yellow on crests and blue in troughs. layers often bigger than those of Urania, some-
times reaching up to 15 air/chitin series. This
ensures high reectance. The phenomenon
Concave Structures: Papilio and here occurs within each basin. A light ray
Cincidelae falling in the middle of a basin is reected and
produces a specic color, in accordance with
Concave structures at the scale of scales are equation (7-8), whereas another falling next to
rather rare. If they generate wonderful col- the middle will be deviated producing color
ored effects, it is more of a shade effect, like slightly shifted towards the blue. There is in-
in velvet for instance, than one of iridescence deed an iridescent phenomenon here, yet the
Papilionidae: Solid/Gas Structures 79

(a) (b)

(c) (d)

Figure 7.20. Structural organization of PapilioPapilio ulysse. (a) cover structural scales. (b) a striated net-
work and undulations forming a counter-striae network. Below, a section performed perpendicularly to
striae next to the center of a basin and showing the (c): SEM, (d): TEM.

polarization effects are minor because of the proceeds from a pointillist effect of the out-
low incidence of the wave. Even further from line, blue framing green at an invisible scale.
the center, the result is totally different. The in- Interferential colors generated by curved struc-
cidence increases, getting closer to the Brew- tures always entail polarization effects by re-
ster angle (Figure 7.22). Polarization effects ex- ection. In the case interesting us here, these
pand and the blue becomes more intense. A effects also occur at the scale of the thin struc-
double reection occurs, making light emerge tures of scales (basins), not always perceivable
in the incident direction. There is no irides- at the buttery scale. The reason why is quite
cence anymore but a mixing of colors with the simple, as illustrated in Figure 7.24.
wavelength coming from the center. Here is a When a nonpolarized wavelength falls onto
peculiar case of combination of structural col- a basin, under normal incidence for instance,
ors. The nal color, with a green predominant, we know that in the areas that are inclined
80 Chapter 7 1-Dimensional Structures: Interferences

Figure 7.21. Iridescent effects on Papilio illuminated under grazing incidence on the right hand-side, and
under normal incidence on the left hand-side. From top to bottom: Papilio ulysse, papilio paris, and papilio
blumei.

Figure 7.22. Reection in a concave multilayer. Under low incidence (a) the center is illuminated under near-
normal incidence and produces a generally yellow-brown nonpolarized color by interference. Incidence is
a bit higher on margins, the color approaches green, goes out under an oblique angle, and remains weakly
polarized. Closer to the basin (b) incidence is close to the Brewster angle, itself close to 45 . The reected
light is blue, highly polarized, and goes out after a double reection, in the direction of incidence. (c) Aspect
of a basin illuminated under normal incidence.
Papilionidae: Solid/Gas Structures 81

Figure 7.23. Polarization ef-


fect on Papilio paris scales. On
the left, the polarization of
the incident wave is paral-
lel to striae. The high edges
of the basin illuminated un-
der high incidence, send back
a blue polarized light in this
direction. On the right, po-
larization is perpendicular to
striae, the small edges of the
basin now reect the polar-
ized wave. Under nonpolar-
ized light, the elongated shape
of basins leads to a reected
light that is slightly polarized
in the direction of striae.

according to the Brewster angle B , only the to it. When the double reection occurs with
s component is reected. If the angle is not ex- light deviating in the incident direction, there
actly equal to B , a small part of the p wave will are two successive reections both increasing
also be reected. Within an angle of 10 around the proportion of wave s over wave p, which
B , the p proportion doesnt exceed 10%. This results in a highly polarized ray, even if the an-
case is particularly interesting, because if the gle is not optimal.
Brewster angle is not exactly equal to /4 However, such a device doesnt create a
more around 55 it remains still very close polarization effect at the scale of the whole

Figure 7.24. Variations of polarizations after two


successive reections under an incidence close to the Figure 7.25. Two cincidelae of the Mediterranean re-
Brewster angle. Emergent waves are very highly po- gions; left, Cincidela hybrida Linne, and right Cincidela
larized, but in perpendicular directions, which gen- campestris Linne. (Photography and collection; Jean-
erally cancels the effect. Paul LeclercParis).
82 Chapter 7 1-Dimensional Structures: Interferences

Figure 7.26. Photon micro-scopy


views of Cincidela hybrida elytrons
(a) and Cincidela campestris, (b).

(a) (b)

buttery, or only indistinctly in some cases. This alveolar system includes small protuber-
Basins present a relatively central symmetry, ances surrounded by some ten alveoles and
which makes that wave s that is reected by presenting a color that can be quite different
the two opposite edges of the basinlet us say from the rest, sometimes violet-blue.
the left and right edgesis a wave p from the The cuticle has a traditional structure, bril-
viewpoint of the two perpendicular edges (top lated in the exocuticle and multilayered in the
and bottom). At the scale of the scales or of epicuticle. There are up to 12 stacked solid lay-
the whole wing, the reected ray includes as ers of 150 to 200 nm. They appear slightly con-
many s polarized waves as p polarized waves. trasted under the electron microscope, hence
The Papilio Ulysse is a good illustration. A dis- a slight contrast in the optical index, probably
crepancy can only occur when certain edges minor yet sufcient to create the resulting col-
are constantly longer than their perpendicular ored effects.
counterparts, like Papilio paris for instance, in All the optical phenomena described in Pa-
which the blue chromatic component is slightly pilios can be found here, especially polarizing
polarized after the double reection. effects in alveoles. However, the latter dont
show any lengthening in a denite direction,
that is why there is no polarization including
Cicindelidae: Solid/Solid Structures the whole insect.

Cicindelidae, or tiger beetles, are small


Coleoptera belonging to the Carabidae with Cetonia: Circular Polarization
green or brown elytrons. The epicuticle surface
consists of adjacent hexagonal alveoles around The last interferential phenomenon to mention
ten microns in diameter. Alveoles present is that associated with optical activity. The Ce-
a hemispheric section, which results in the tonia family (chafers) is well known for its ex-
same phenomenon observed in Papilios at a traordinary insects with bright metallic glance.
bigger scale and with more diverse colors. It also commonly shows helicoidal structures

Figure 7.27. Polarization effect re-


sulting from reection in the basins
of Cincidela hybrida elytrons. On the
left, the polarization direction is ver-
tical, while it is horizontal on the
right. The yellow background is
not polarized contrary to the green
edges of basins.
Cetonia: Circular Polarization 83

(a) (b)

(c)

Figure 7.28. SEM images of volume and surface structure of Cincidela campestris. (a) Surface alve-
olae and protuberances. (b) A section of the cuticle showing the exocuticle chitineous brillae
and the nonbrillated epicuticle. (c) The multilayer hemispheric structure of the epicuticle in an
alveola.

that generate the peculiar polarization effects, This case is an anisotropic solid/solid mul-
as we have already described in the previous tilayer structure with a director rotating from
chapters. one layer to the other, creating a periodic gra-
dient of index, though quite weak, in the epi-
cuticle. This little difference in index is com-
pensated by the large number of layers leading
to the famous bright glance. It is important to
note that there is no alternation of materials
with low and high indices in the multilayer.
Yet, the latter results from the rotation in its
plane of a layer composed of a single mate-
rial with two different indices in perpendicular
directions. Interference occurs between layers
with the same direction. It is associated with
Figure 7.29. The Rose chafer Cetonia aurata in the an optical activity, which can be demonstrated
ying position. by using a /4 slide introducing a delay of
84 Chapter 7 1-Dimensional Structures: Interferences

Figure 7.30. Two Hoplia of Eu-


rope. Left, Hoplia argentea and
right, Hoplia cerulea.

130 to 135 nmso, a /4 slide for a wavelength The Scarabeidae family includes other mem-
= 550 nm corresponding to the green color of bers with elytrons covered with scales. The
the Rose chafer for instancefollowed by a lin- arrangement, morphologies, and structures of
ear polarizer. these scales differ greatly from one species to
All the measurements suggest that polariza- another. We have already mentioned in the
tion is counterclockwise on the two elytrons third chapter the singular arrangement of Ho-
and that the axial symmetry of the insect is plia cerulea scales, resembling that of a sh, but
thus not respected. The origin of the layer ro- also as singular the scales of Hoplia argentea that
tation remains obscure. It could be the result are covered with thin radicles.
of the daily alternation of illumination dur- In the absolute, treating such a problem is
ing the nymphal stage. It can be suppressed tedious, yet it is somewhat simplied in the
by keeping the insect exposed to a constant present case, in which the coefcient of phase
illumination. delay = (ne n0 )k0 is large compared to the

(a) (b)

Figure 7.31. Cross section of Hoplia scales. On the left, a female Hoplia argentea scale and Hoplia cerulea
on the right. Layers are well visible and set equally apart, but their individual structures are quite
disorganized.
Cetonia: Circular Polarization 85

rotation coefcient of the director given by angle and the one remaining parallel to the
= Z. (ne and n0 are the extraordinary and quick axis (Ex on the gure) is ahead with a
ordinary indices of the material, k0 the wave phase difference of .
vector, and Z the traveled distance measured
from the surface.)
One thus demonstrates that the Jones matrix
of the system giving the polarization state of
the wave after a travel on a distance d can be
expressed as follows:
 
d
exp j 0
2
T = R(d)   ,
d A microscopic observation reveals in the
0 exp j scales a multilayer solid/gas-type structure
2
without any specic orientation of brilles. Ho-
i.e., like the product of an optical activity and plia argentea radicles, whether male or female,
a dephasor. 300 nm in diameter and 1 to 2-m long, scatter
After a travel on a distance d in such a ma- the reected light and weaken substantially the
terial, the two components have rotated of an elytron brightness.
8
2-Dimensional Structures:
Interferences and Diffraction

Theory Recalls If one considers an incident monochromatic


wave under an angle i on two groves spaced
Diffraction by Gratings of a, the difference of period between two
diffracted waves in one direction i is:
It has been noted that buttery scales showed  
on their superior faces a system of striae, of- = a sin i + sin i . (8-1)
ten very regular, which can explain irides-
cence through diffraction. Many biologists in The interference between the two waves is
the late 20th century defended this hypothesis constructive when they are in phase, i.e., when
and Michelson again was attracted to it for a the period difference d is a whole multiple of
moment. We will briey review the principle the wavelength:
 
of grating and the most obvious characteristics a sin i + sin i = k, (8-2)
of the color created through diffraction.
A plane grating is a system consisting of nu- or also:
merous diffracting elements, known as groves, sin i + sin i = kn (8-3)
and regularly spaced. Groves can be slits sepa-
rated by opaque areas in transmission gratings, where n is the number of groves per unit of
or mirrors in reection gratings. As with in- length.
sects, iridescence is the more often observed in If a white light illuminates the grating, one
reection, so we will illustrate our study with concludes that for k = 0, all the wavelength
the latter, the physics of which is the same of interfere constructively in the same direction
the former. (the direction of the specular reection i =
When an electromagnetic wave falls on such i ), thus producing white light. But one also
a grating, it is diffracted in all directions by each concludes that the wavelengths are separated
grove serving as secondary sources, so that out- and form spectra for all of the other orders (k
going waveshere the reected waveswill = +/ 1, +/ 2 . . . ).
interfere. For a given wavelength, these waves For a grating of 1000 groves per mm such as
will be in phase in certain directions and they one formed by the scale striae, and a blue wave-
give constructive interferences, while they will length = 400 nm under normal incidence
be out of phase in other directions leading to (i = 0), equation (8-3) shows that k can only
destructive interferences. These directions thus be equal to 0, 1 and 2 with corresponding
only depend on the grating period and the inci- deviation angles around 0, 23 30 and 53 6 . In
dent angle. Under white light, each wavelength the red, at the other end of the spectrum ( =
will interfere constructively in a different di- 700 nm), k can only be equal to 0 and 1 and the
rection. These two aspects are characteristics deviation angle is around 44 30 . The higher k
of iridescence. and the larger the wavelength, the greater the

86
Theory Recalls 87

Figure 8.2. The optical path difference between two


beams diffracted by two adjacent groves depends
on the angle of incidence i, the angle of diffraction
i , and the grating period a. The larger the period
a and the difference between i and i, the higher
the phase difference between the two waves.

The rigor that is necessary to address such


aspects would go beyond the limits of this
book. The structures found among Morphos and
their optical properties, which will be stud-
ied later, are sufcient to convey the richness
and complexity of the phenomena. We can still
mention them by wondering why, for instance,

Figure 8.1. One of the most widespread examples of


a grating; diffraction resulting from burning a com-
pact disc.

dispersion. Contrary to prism, and in every or-


der, red is more deviated than blue.
I described in the most simple way the prin-
ciple of diffraction and the properties of the
diffracted wave. Many aspects of the theory
have purposely been avoided, yet what has
been mentioned allows understanding the phe-
nomena observed among certain butteries.
Yet, other phenomena cannot be explained
as simply, especially those related to the po- Figure 8.3. The number of orders is limited by the
larization of the diffracted waves. We wont sine modulus ranging from 0 to 1. In the illustrated
case, the period is 1000 groves/mm, which is close
either address complex gratings, multiple- to that of iridescent butteries. Only the orders with
periodicity ones for instance, whereas they k = 1 are whole. When k = 2, only the shortest
characterize those found among butteries. wavelengthsgreen and blueremain visible.
88 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

is required to produce the second effect. Inter-


ferences, a phenomenon selecting short wave-
lengths, also occur within each of the striae.
They are constructive for blue waves and de-
structive for reds. As we will now see, striae
indeed show a vertical lamellate structure.
This conrms the double periodicity in
two different directions: the rst in striae in
the plane of scales (and slightly wings), and
the second in the striae lamellae perpendic-
ularly to the scales. Here lies all the com-
plexity of Morphos, and the origin of their
beauty.
Gratings are commonly used in many lab-
oratory materials. In spectrophotometers, for
instance, they are used to separate the vari-
Figure 8.4. The different orders. In order 1, two ous wavelengths of a light beam, but also in
waves proceeding from two neighboring motifs many other applications, like CD players. Ev-
present an optical path difference of one wavelength,
three in order 3 . . .
eryone has at least once looked at the iridescent
colors of light rays reected on a microgroove
and even more on compact disks possessing
the whole optical spectrum is never observed a higher reection factor. They are yet much
in the light as reected by Morphos wings. The rarer in nature, as the detailed study of Mor-
latter obviously present a grating structure but phos will demonstrate. Diffraction partly gen-
its color varies between only violet and blue- erates the colors of the coleopter Serica sericea,
green. In the traditional presentation that has and of certain arachnids like Cosmophasis tha-
just been made, we have assumed that each lassina. Yet, it is often difcult to distinguish the
grove interacts with each wavelength in a sim- respective roles of interference and diffraction
ilar way. Although the term is not appropri- in the colors resulting from these structures, as
ate for such tiny elements, it has also been as- they tend to be interwoven. In the case of sur-
sumed that the reection coefcient of the mir- face gratings, a rather simple technique is to
rors forming groves was the same along the take a replica of the surface and to examine its
whole spectrum. Yet, one can assume that it is color. In this way, one gets rid of interferential
different and that in one way or another, only structures that are 3-dimensional, and only the
short wavelength are diffracted, the remaining colors generated by gratings remain.
ones being transmitted and absorbed by bot- Let us nally note that contrary to inter-
tom layers. Only reected colors will be found ferences, diffraction produces color, but also
in every order, including in the 0 order, which scatters it spatially. In this way, this function
is not white anymore. doesnt depend on scales anymore, which are
Another possible cause would be the pres- usually quite at, as we are going to see now.
ence of a period so short that no whole order
can be visible and that only the smallest wave-
lengths of the rst order, which are the less de-
Type Butteries with 2-Dimensional
viated, appear. Let us note that in this case, the
whole of the spectrum should be in the 0 order,
Structures
thus appearing white.
Morphidae
The reality is more complex of course, and
is somewhere between the two assumptions! Morphidae is slightly more complex. The Mor-
In fact, the periods found on scales are quite pho genus (there are two others in the fam-
close, among certain species, to 0,5 m, which ily: Antirrhea and Caerois) includes more than
Type Butteries with 2-Dimensional Structures 89

Figure 8.5. (a) Selective diffraction. Long wavelengths are either transmitted or absorbed. Only short
ones can be found in each order, including in order 0. (b) High dispersion. Under normal incidence, a
2000 groves/mm grating diffracts only the wavelengths shorter than 500 nmblue-greenin the rst
order, while they are all present in the 0 order.

30 species and many sub-species, all present- a priori. It is also a Morphalidae widespread in
ing a strong sexual dimorphism. That is why it the Amazonian basin and even in Bolivia. The
is not an easy task to choose a species for our male is entirely blue on its dorsal side, except-
study. Our predecessors tended to choose one ing the abdominal and costal edges, fringed
more because of some opportunity rather than with black. It is cryptic on its ventral side
based on rigorous scientic reasons. with three ocelli on posterior wings and two
All male Morphos show marked iridescent ef- on the anterior, all little developed. As has al-
fects with a predominant blue. Some are very ready been shown in Chapter 5, these colors
dark, of a deep blue, while others are almost vary with the exterior medium index, demon-
white with only a slightly bluish touch. Many strating without any ambiguity their structural
present marked bright colors, whereas others
have a matte surface that seems to extensively
diffuse light. One could thus classify them very
basically in a plane according to two criteria:
brightness being the rst coordinate and rang-
ing from 1 for a mirror to 0 for a perfectly
scattering surface, while the second coordinate
representing spectral purity ranging from 1 for
a monochromatic color to 0 for white. This is
merely a subjective classication, which isnt
based on any measurements. We have chosen
our standard Morpho (Morpho menelaus) among
the species classied in the center of this plane,
eventually selecting the more common ones in
order to study them morphologically and opti-
cally, and to widen our study with some other
species of the plane by focusing more on dif-
ferences.

Morpho Menelaus: Characterization and


Modeling Figure 8.6. Iridescence curve of a Morpho menelaus in
the CIE chromatic diagram from normal incidence
Our choice proved relevant, Morpho menelaus to grazing incidence. The curves are interpreted in
revealing average in every way as it seemed chapter 13.
90 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

Figure 8.7. Reection factor of a Morpho menelaus whole wing for different negative incidences, (from the
base to the apex). Angles have been measured from the normal to the wing plane. One must add 10.5 in
order to obtain incidence on the interferential structure.

origin. Immersed into a medium with a rather tion, amplitudes differ, as can be seen on the
high index, close to that of chitin, and when col- chromatic lines obtained from spectral mea-
ors have totally vanished, the buttery looks surements and which, contrary to an isotropic
extremely dark, revealing concentrated pig- thin layer, do not overlap. This observation
ments on either face. questioning Michelsons interferential hypoth-
Iridescence is particularly marked, as shown esis could, on the contrary, conrm it. As has
by the series of pictures in Figure 8.8. The been seen in the last chapter, the reection min-
evolution and characteristics of colors will be ima of an electromagnetic wavelength on a thin
studied further, but let us now rst focus layer are given by:
on the aspect of the buttery under various
kmin = 2ne cos t
illuminations. What is the most remarkable
initially is the important dissymmetry of col- This equation can be rewritten by involving,
ors in the wings when illuminated under the not the t refracting angle, which we dont a
same angle from the right or the left. This sug- priori know, but the i incident angle. Snells
gests that reecting structures on the wings law allows to modify this variable and results
are inclined. An obvious fact that Michel- in the following apparently more complex ex-
son missed, yet he had then to assume that pression:
the structure existed. Visible and near-infrared
spectrometric measurements, performed on k2 2
sin2 i = + n2 (8-4)
the big homogeneous areas of cells M3, Cu1, 4e2 m
Cu2, and discoidal, conrm this initial visual Nevertheless, under this form, one notes that
impression. sin2 i varies in a linear way with the square of
Indeed, through measurements performed the wavelength m. The slope of this straight
from the apex towards the base or the re- line gives the thickness of the layer and the ab-
verse, one can observe a shifting in the spec- scissa at the origin its index. In pre-computer
tral structures towards blue when the incident times, this simple graphic method allowed us
angle increases, with maxima and minima in to determine indices and thickness layer. Sub-
various places for a given incidence. In addi- jected to spectral measurements performed on
Type Butteries with 2-Dimensional Structures 91

Figure 8.8. Iridescence effects on Morpho godarti exposed to collimated illumination. One can note that
colors vary quite substantially depending on the illumination directionfrom base to apex or the
opposite.

M. menelaus, one notes that, like Michelson, The dissymmetrical colors made us conclude
this linear variation isnt respected, and by far. that the structure was inclined. It is therefore
Experimental points form large arcs of circle: necessary to correct every incidence angle of a
concave for positive incidences and convex for constant angle . Once done, all the points of
negative ones. the two arcs of circle converge towards their
92 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

Figure 8.9. Positive inci-


dences (apexbase) Negative
incidences (baseapex)
Wavelengths of minima of
reection of Morpho menelaus
wings in relation to the
incidence measured from the
normal to the wing plane
(--- r---) and after correcting
the angle s of the structures
on the plane ( ). The best
alignment is obtained when s
is equal to 10.5 . The thickness
of the corresponding layer
leading to these interferences,
given by the slope of the
straight line, is e = 0.186 m
and its index n = 1.68.

common chord, giving thus the index, thick- tivity, as has been seen in higher proportions
ness, and angle of the reecting structure, here among C. madagascariensis.
about 10 . Measurements performed in the vis-
ible and the near infrared on zones of complete
wings or on wings deprived of the scales of the Structures of Wings and Scales
dorsal side, and then of the ventral, allow one
to underline the inuence on the total reec- Let us discover this structure by zooming in
tion of each layer. Dorsal scales are responsi- by six orders of magnitude. By using optical
ble for the visible structure of spectra and thus microscopy, one observes that scales are here
color, whereas the alary membrane and pig- again arranged in two layers, yet structural
mentary scales are in charge of infrared reec- ones are in the bottom. Cover scales are thus

Reflectivity

0.8

0.6 Wing Without dorsal scales

0.4

0.2 Figure 8.10. Diffuse reection factor un-


Alary membrane der normal incidence of a whole wing of
Morpho menelaus, then of the wing from
250 500 750 1000 1500 2000 which the dorsal scales were removed,
Wavelength (nm) and nally of the only membrane.
Structures of Wings and Scales 93

Figure 8.11. Morpho menelaus in


the open air and under index
liquid. Colors vary similarly to
Morpho godarti, yet the layer of
the dorsal pigmentary scales and
to a lesser extent, the pigments
of the dorsal structural scales are
highly absorbing, and the butter-
y turns almost black when im-
mersed in a liquid of high refrac-
tion index.

deprived of pigment so that light can come and disappear completely and one can note that
play on the structures of the ground scales that ground scales, if they have lost their colors, still
are partly covered. Immersed into a liquid of serve as a dark screen. From different origins
adequate optical index (Figure 3.22), the latter among C. madagascariensis, the two structural
and pigmentary effects here on the contrary
proceed from the same scales. The inuence
of cover scales is minor and it is possible to
observe them by using reectance optical mi-
croscopy only under grazing incidence, where
a slight scattering effects makes them appear
whitish by diminishing the brightness of cov-
ered areas.
Scanning electron microscopy reveals the
same dense arrangement of structural scales
that form a regular background all over the
wing almost without overlapping. Structural
Figure 8.12. Optical microscope view in reection scales are at, rectangular (80 m 200 m)
of Morpho menelaus dorsal scales. Long and narrow, and slightly rounded and fringed at the apex.
cover scales are almost invisible in transmission or The cupule axis from which scales issue, is
in reection under normal incidence. Immersed in strongly inclined on the wing plane, thus form-
an index liquidFigure 3.22structural scales look ing a layer parallel to the alary membrane.
brown. They contain pigments and take part along
with ventral scales, in the formation of an opaque Striae are also dense (1600 to 1800 per mm)
screen that absorbs the wavelengths that are not re- and seem very deepsome even tend to join
ected by the structure. in places. Cover scales are slightly longer and
94 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

(a)

(b)

Figure 8.14. Morpho menelaus. SEM images of a cover


scale, (a) and of a structural ground scale, (b).
Figure 8.13. (a) Photonic microscope image of
M. menelaus cover scales observed in reection. The
reected color is extremely rarely visible. Its inten-
sity is weak and very directional. (b) Striae and inter- lar to ground ones from a structural point of
stria space of a M. menelaus cover scale. The number view. However, they differ on three aspects,
of lamella is limited and the period is large. which undermine their effects on the wing
color (1). They arethis is true for all the
narrower (50 m 250 m) and their striae species that are observeddeprived of pig-
much more spaced (around 2 m) and linked ments. Non-reected wavelengths are trans-
together through some bridges. If one exam- mitted to the ground scales on which they
ines striae more closely, the two types of scales lie (2). Spacing between striae is much more
still appear quite different at this scale. The important. Striae form a scarcely dispersing
lamellae of structural striae overlap in a large grating with a period of 2 to 4 m (3). The
part, the ratio between their spacing and length number of lamellae is limited, four in aver-
sometimes reaching up to 10 or 11. This leads age among M. menelaus; this substantially re-
to a structure of around 20 to 22 alternated duces the general reection factor of the scales.
air/chitin layers, every lamella looking slightly Cover scales actually hardly contribute to the
plane when covered, but becoming cylindri- wing color and to spatial dispersion of light.
cal at their ends. Finally, they are inclined on This function is fullled by ground scales,
the plane of scales under an angle of about whereas cover scales let light enter them with-
10 , as it could be assumed with the adjusted out hindering it in any way. As will be seen
arrangement of minima. Cover scales are simi- later, it is not always the case. Nevertheless,
Structures of Wings and Scales 95

Figure 8.15. Morpho menelaus. On this image, one can


distinguish between two striae of a structural scale
up to 8 overlapping lamellae linked by a tight net-
work of vertical microtrabeculae.

in order to play a critical role in coloration,


their arrangement and structure must be
modied.
Observed by using transmission electron mi-
croscopy, a transversal cross-section of these
scales shows the real shape and arrangement of
lamellae at last, which are alternately arranged
on both sides of the stria median axis.
The modeling of optical properties of M.
menelaus proves much more complex than that
of Uranidae. Scales are certainly at and almost
parallel to the alary membrane, contrary to the
latter. Yet, two aspects make it more compli-
cated:
In relation to the longitudinal axis of scales
on the one hand, lamellae, with a nite length, Figure 8.16. (a) Above, a cross section of striae show-
are tilted on the membrane plane from which ing the arrangement in staggered rows of lamellae
on both sides of the central axis. (b) Below, on the lat-
they spring. Therefore, it is not merely a multi- eral view of the same striae one can note the angle
layer that is innite in every direction and, the formed by lamellae and the scale membrane, around
inclination of which would have been modi- 10 to 15 .
ed, but in reality a supported and truncated
multilayer.
According to a transversal axis on the other width of lamellae, l, is constant in most cases.
hand, the stacks of lamellae are translated to The distance d between the edges of two neigh-
one another by a half-period. boring lamellae can also be considered as con-
Between the two obstacles, the rst is the stant. Vertically, each stria consists of alternat-
most obstructing; yet they dont really inu- ing air/chitin layers, the thickness of which
ence the physics of the phenomenon. The incli- slightly varies along the stria averaging 100 nm.
nation and translation of lamellae can therefore Under near-normal incidence, each stria be-
be neglected and let us gure in the most basic haves like the grove of a grating with a high
way each stria as a horizontal overlapping of reection/reectance/reecting coefcient in
lamellae presenting rectangular sections. The blue, while the whole forms a selective grating
96 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

Figure 8.17. The two schematiza-


tions for the structure modeling.
Longitudinally, the stack of lamel-
lae (a) is assimilated to a tilted in-
nite multilayer (b). Transversally,
the two edges are shifted from one
another by the thickness of one
lamella. The reected rays on one
side or on the other are thus not
systematically in phase.

composed of 1600 lines per mm on average, ite materials and the denition of the index of
thus only including three orders (1, 0, +1) the components requires a rather complicated
in which, as has been mentioned, the smaller modeling, which wont be fully presented
wavelength are less deviated than larger ones. here, as well as a theory of effective medium.
More generally, and if striae are very dense Maxwell Garnetts theory as mentioned in
(d  e), the whole behaves like a quasi- Chapter 6, with a depolarizing coefcient A
continuous multilayer generating thin layer in- corresponding to the elongated shape of inclu-
terferences. And when one exposes the insect sions, is well adapted to the present problem.
to a diffuse light (falling under all incidences), It leads to an anisotropic solid phase index
more short wavelengths will be observed un- chitin and proteinwhich means that it shows
der a high angle of incidence than large ones. a different value depending on the direction,
The two phenomena are antagonist and unless that of striae, or the perpendicular one. It is in-
one is predominant, this results in color mixing teresting to note that the two index curves inter-
and thus to a reduced spectral purity. sect in the 450 nm500 nm spectral range; that
Despite the good atness of the Morpho is, the blue-green part of the spectrum that cor-
menelaus scales, their orientations prove too an- responds to the maximal reection of Morphos.
archical and the wing distortions are such that This certainly explains why researchers have
it is illusory to expect to observe the diffraction so far missed this anisotropy, as they used to
orders from the entire wing. It then becomes dene indices through immersion techniques.
possible to represent the wing of Morphos as a As has already been described, the method con-
stack of alternated layers of low and high in- sists in dipping the wing into diverse index liq-
dices. High-index layers that contain lamellae uids and seeking the perfect match revealed
mostly consisting of chitin with a few air, while by the disappearance of colors. This merely
low-index ones are quasi-symmetrical. These matches the index with the wavelength of the
layers are thus basically composed of compos- color. In the present case, it is isotropic.

Figure 8.18. When the spacing be-


tween striae decreases, the diffrac-
tion effect is restricted to low-
incidence rays and is supplanted
by a multilayer interferential pro-
cess. It is the case with Morpho
menelaus where striae are almost
adjacent and the optical properties
of which can be determined thanks
to a simple multilayer model. In
both cases, one can take into ac-
count the effect of air inclusions or
of chitin trabeculae on the layer in-
dices by using an effective medium
theory.
Structures of Wings and Scales 97

Figure 8.19. The three stages in the modeling of Morpho structures by using an effective medium theory.
(a) Denition of a schematized structure model. (b) Determination of effective heterogeneous lms. (c) Ho-
mogenizing of lms by using the effective medium theory.

Figure 8.20. Modeling of the reection of Morpho menelaus wing under normal incidence by a stack of
14 alternated air/chitin layers. Thicknesses are in fairly good agreement with those deduced from TEM
observations.
Cuticular index

Transverse Electric

Transverse Magnetic
1.5

1
Figure 8.21. Opposite. Variations of the
400 450 500 550 600 650 700 750
anisotropic index of the cuticle scales de-
duced from the model (see Figure 8.19). Wavelength (nm)
98 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

essence. That is why they are so frequently


used in laboratories or in industry. They are
very efcient in Morphos and enough to largely
disperse light. One actually notes that most
of the scales of Morphos are perfectly at. We
have established a diagram of the light inten-
sity diffracted by one of the brightest Mor-
phos (Morphos zephyritis) at its reection max-
imum of the spectrum, in the blue, at =
450 nm. It is meaningful. Falling onto the axis
Figure 8.22. Diffraction chart of Morpho zephyritis. of the object, under a relatively low incidence
The insect is illuminated under an incidence of 20 in of 20 , light is dispersed laterally on more than
the normal plane = 20 ,  = 0 . Light is scattered 100 , and only of 20 in the plane of incidence.
laterally over more than 100 , but only over some 20 This remarkable arrangement, associated with
in the incident plane. a high reection factor, provides the buttery
with an excellent visibility in the full space,
Dispersion and Polarization which is essential for Morphos. Indeed, sexual
dimorphism is strong in this category. Females
Contrary to interferential structures, diffract- are cryptic, but sexes are also separated geo-
ing/diffractive systems are dispersive in graphically, the latter living around the canopy,
whereas males can be found in the rst strata
of the forest. Their recognition is essentially vi-
sual, hence the importance of this high visi-
bility, even if it can also attract predators . . . or
hunters.
We will now tackle intraspecic communica-
tion and focus more on differences than simi-
larities, because it is as crucial for Morphos to be
able to distinguish themselves as to recognize
between males and females. These differences
lie in the relative arrangement of cover scales
and their thin structures. Before making an
overview of the main species, let us review the
blue Morphos menelaus and the light Morphos
godarti, the chromatic variations of which have
been described in the previous plates.
From the most to intense blue to the purest white,
from the brightest to the matest
Even if the modeling of the optical proper-
ties of Morpho menelaus wings is still to be per-
fected, this buttery allowed us to evaluate the
numerous phenomena that play a part in its
colors. Our knowledge is certainly incomplete
concerning the precise place of scales, and their
indexespecially in ultraviolet and infrared
as well as the quality and arrangement of pos-
sible pigments. Yet our present knowledge is
Figure 8.23. Morpho menelaus striae above and
M. godarti below. Structures are strictly similar. precise enough to be used as a reference and
M. godarti can sometimes present even more over- guide the way for research. We dont know
lapping layers, but its blue color looks much paler. if one could develop a convex micrograting
Dispersion and Polarization 99

Figure 8.24. The basic structure of iridescent butteries. First, a scattering layer formed by cover scales.
Its efciency depends on the covering ratio ranging from 0 in M. cypris to 80% among M. godarti. Urania
are deprived of scattering layer. Second, a reecting layer formed by structural scales, which are cover
scales among Urania and ground scales among Morphos. Its efciency depends on the number of layers, the
spacing between striae, and the scale shape. Finally, an absorbing layer formed by ground scales, the alary
membrane, and the ventral side. Its absorption factor depends on the pigment concentration but also on
the structure of pigmentary scales.

Figure 8.26. In most species, the covering ratio


Figure 8.25. The covering ratio of ground scales by ranges from 80% to 50%, exposing ground scales to
cover scales varies with the species. It is almost 100% a large extent. (above: M. anaxibia, center: M. didius,
among M. helenor and M. adonis. below: M. godarti.
100 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

(a)

Figure 8.27. Among certain Morpho, cover scales are


very atrophied and almost invisible. (above: M. aega,
center: M. rhetenov, below: M. cypris)

engraved in a Bragg mirror, or even if it is


worth it; yet we know that it exists and that
its properties can be dened. Let us begin by
studying the variations of predominant colors.
Morpho menelaus is a rather deep-blue buttery. (b)
Its predominant wavelength varies with inci-
Figure 8.28. (a) M. zephyritis scales immersed in an
dence from the blue-green 495 nm) to the index liquid, n = 1,478. Cover scales have completely
violet ((m 460 nm) with a maximal purity of vanished and do not include any pigment. Among
40% in the blue ((m 480 nm) for an incidence M. Sulkowskyi, (b) ground scales do not contain any
close to 30 . If Morpho can hardly be bluer, it is pigment either and they also vanish when immersed
often whiter. M. polyphenus breaks the record in an index liquid.
as it appears totally white. However, it is quite
rare, so we chose a more widespread one that
is less white: the Bolivian Morpho, M. godarti, screen here, which makes interferential colors
which enabled us to illustrate how the medium appear. Still through photonic microscopy, or
index inuences structural colors. SEM at low magnication, cover scales appear
This experiment gives us a rst key to de- at and slightly scattering. They are shorter, yet
cipher this evolution towards white; that is, much wider than those of M. menelaus, thus al-
a decreasing purity. In an index liquid close most entirely concealing structural scales, the
to that of chitin, wings look almost transpar- covering ratio approximating 80%, while it is
ent, exposing the ocelli on the ventral side. The only 40% among M. menelaus. It is in these two
wing contains very few pigments and the same aspectslack of a dark screen, and occurrence
experiment, by using photonic microscopy, al- of a scattering layer on the frontal facethat
lows us to locate them in the ventral side scales lies the main cause of the loss of spectral purity
only, dorsal ones becoming totally invisible. among M. godarti. A decrease that is partly due
Contrary to Morpho menelaus, there is no dark to the arrangement and structure of striae. The
Dispersion and Polarization 101

Figure 8.30. The structure of M. zephyritis scales,


above, and M. Sulkowskyi, below. Cover scales on
Figure 8.29. Among the two species, M. zephyritis the left and ground scales on the right. Structures
aboveand M. Sulkowskyi, the two scale types are are almost similar.
juxtaposed and iridescent.

spaces separating the latter are slightly wider the two membranes of the wing develop in-
(1200 striae per mm on average instead of 1800 dependently at the nymphal stage, so that the
among M. menelaus). The multilayer effect is motifs on the two faces of scales of different
diminished and more intertwined with the an- structures do not show any similarity. It is the
tagonist effect of the grating. Iridescence is less case of M. cypris that presents, like most of
marked, and more due to variations of reected the other members of the genus, cryptic col-
intensity than to real color changes. ors ranging from brown to black on the ven-
Let us now mention brightness variations. tral side, precisely excluding the dorsal white
Among the numerous very bright Morphos, in- spots. This perfect correspondence doesnt pro-
cluding M. aega as studied by Michelson, we ceed from the simultaneous development of
selected Morphos cypris. This buttery is char- the faces, but from the relation of cause and
acterized by long white streaks on the medial effect.
areas of its wings. These streaks rigorously Concerning the brightness of M. cypris, high
conrm the conclusions of our previous study magnications are not necessary to nd its
and we wont elaborate on them. This butter- cause. Photonic microscopy instantaneously
y highlights particularly well the inuence offers it. Cover scales are completely atrophied
of the absorbing sub-layer. As has been seen, and expose the total surface of structural scales
102 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

Figure 8.31. Polarization effect of the incident wave


on the color of M. zephyritis. On the right, the in- Figure 8.32. Diffracted elds as calculated from a
cident wave is polarized in the Transverse Electric schematized structure of striae. The incident wave
Mode, whereas it is Transverse Magnetic on the left is not represented here. Its incidence is normal in
picture. both cases. The TE wave diffracts in the R1 order
on the rightand the TM wave in the R1 order, on
the left.
to light. Like those of M. menelaus, the latter are
also slightly absorbing and very at. Besides,
their striae are particularly dense (more than the properties of cover scales, but enable them
2000 per mm) and composed of a great number to perfectly adapt to structural ones. As for
of layers (between 10 and 12 lamellae). All these the latter, the curve of the mirror makes wave-
elements further the mirror effect. lengths that interfere in a given direction mix,
If there are many a bright Morphos, there hence leading to a slight iridescence. Only a
is but one that is matte, the Morpho anaxibia, few species among an approximate total of
which will enable us to complete our demon- 30 composing the genus have thus been de-
stration. It is little iridescent and is complemen- scribed. However, the remaining will be clas-
tary to M. cypris in all of its aspects. Cover sied according to the four poles revealing the
scales form a quasi-continuous scattering front variations observed, which will be illustrated
layer. However, the principal characteristic of in the following plates. Thanks to them, we
the buttery lies in the fact that the two kinds have at least been able to dene the evolu-
of scales are convex. This doesnt affect much tion of colors and brightness among Morphos
Dispersion and Polarization 103

R
I
Ordre R1
0.3
TE
R1

0.2 R-1
R0

TM 30 20 10 0 10 20 30
0.1 Facteur de reflexion (%)

400 500 600 (nm)

R
Ordre 0 (2)

TE
TM
400 500 600 (nm)

Ordre R-1 I

0.2 R1

TE
R-1
0.1 R0
30 20 10 0 10 20 30
Facteur de reflexion (%)
TM
400 500 600 (nm)

Figure 8.34. The reected spectra in the different or-


ders calculated from the same schematized struc-
ture, showing the chromatic shift between the R1
and R1 orders and their various polarization states.
The central order R0 is very weak and hardly polar-
ized. The inset indicates the diffraction direction of
the order maximum.

Figure 8.33. Reection factor. The same effect as the terferences or diffraction, wavelengths devi-
previous, yet it is less distinct among Morpho godarti ated forward are selected. Here is the origin
asarpaTE on the right, TM, left. of color and iridescence.

or urania, which must also occur among oth-


ers with slight differences. Three elements con-
tribute to the production of color in general
among iridescent butteries. They can some-
times be combined in the same structure:

1. An absorbing sub-layer that can either con-


sist of scales of the ventral side, ground
scales of the dorsal sidewhether structural
(like among Morpho) or not (like Urania)or
nally of the alary membrane itself. It in-
cludes dark pigments, mostly melanin. Ab-
sorption is amplied by the alveolar struc-
Figure 8.35. Reection photonic microscope image
ture trapping light. of a male Morpho adonis cover scales, on the right,
2. A partially reecting structure composing and ground scales on the left. Cover scales overlap
the intermediary layer, in which through in- on several layers.
104 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

(a)

Figure 8.37. The bright Morpho cypris and Morpho


rhetenor and the very mat Morpho anaxibia. The re-
ecting structures of the former are inclined to a
(b) large extent on the wing plane and the specular com-
ponent of the reection is signicant, so that under
collimated light, one cannot see color on both wings
at the same time. M. anaxibia presents convex and
very numerous cover scales, whereas the former are
deprived of them.

3. A possible scattering front-layer that more or


less covers the intermediary layer and alters
the brightness of the color resulting from the
latter.

Evolution of the Structures


(c)
This great diversity is also encountered at a mi-
Figure 8.36. SEM image of Morpho adonis. (a) some croscopic scale in the thin structures of each
large cover scales have been removed in order to scale. Among the few species that have been
expose ground scales. In the (b) picture, the very presented and actually among most of others,
regular striated network of cover scales. (c) a lateral
image showing the single-lamella structure. cover scales are scattering as has been seen
(except for M. adonis, perhaps) and brightness
Polarization Effects 105

Figure 8.38. (a) and (b) The various Morpho species examined. The genre includes some 30 species, most of
which are iridescent. Yet, the micro structural study of the latter is incomplete.

can be modied by affecting their covering Polarization Effects


ratio. However, more original solutions based
on the structures of these scales this time exist. The polarization effects generated by complex
Among two species at least, M. sulkowskyi and 2-dimensional structures still remain somehow
M. zephyritis, cover scales become structural obscure and it is difcult to locate some of them
and take part in the coloration of the insect. experimentally. In this way, calculations will
There is no scattering layer on the surface and show the way.
the two butteries show an extremely bright These effects are rst colorimetric. Reec-
aspect. The distinction between cover/ground tion spectra are shifted from one another de-
scales becomes meaningless here since scales pending on whether the incident polarized
are arranged side by side. Surface structures wavelength is Transverse Electric or Transverse
that are very dense, are almost identical. It Magnetic. This is easily perceivable by using
seems that cover scales tend to be deprived photonic microscopy, and the effect seems to
of pigments and that they vanish once im- be systematic, yet with variable intensity de-
mersed into a index liquid. This is not a per- pending on the species. A buttery illumi-
fect test though, because a certain number of nated by a Transverse Electric (TE) polarized
ground scales are also deprived of pigments wave always appears bluer than the same one
(M. sulkowskyi). illuminated by a Transverse Magnetic (TM)
106 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

(a) (b)

Figure 8.39. Colias crocea in the visible (a) and in ultraviolet (b). The wings of the female uniformly absorb
in the ultraviolet spectrum, whereas the male posterior wings reect it partially.

Figure 8.40. Diffuse reection fac-


tor of male and female Colias cro-
cea anterior and posterior wings. An-
terior wings behave strictly simi-
larly throughout the studied range,
whereas posterior ones behave dif-
ferently in ultraviolet. While male
ones are weakly reecting, female
ones totally absorb these radiations.
Polarization Effects 107

Figure 8.41. Diffuse reection factor


of male and female Gonepterix cleopa-
tra anterior wings (5 cm). Male an-
terior wings show reection in ultra-
violet, contrary to its posterior wings
and to female wings.

Figure 8.42. SEM image of Gonepterix cleopatra Anterior wings. (a) General view showing the two scale
types, very distinct here. (b) Multilayer structure of cover scales. They are responsible for iridescence in
ultraviolet. (c) Ground scales contain numerous pterinosomes and generate the orange-yellow color in the
insect.
108 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

Figure 8.43. The same arrangement among females (a) yet structures are identical in both types of scales
(b) and (c). Lamellae are short and do not overlap and both scales contain pterinosomes.

wave, making it turn green. This corresponds light, the wavelength is diffracted on one side
to theoretical assumptions, yet it goes further. and in TM polarized light on the other side.
In fact, this diffracting structure polarizes light, Although it is accepted that Morphos distin-
even (and especially) under near-normal inci- guish the two polarization states, the advan-
dence, which is quite singular! In other words, tage of the phenomenon for them remains un-
diffraction orders are polarized in very dif- clear, so is its origin. One can at least argue
ferent ways when the structure is exposed to that it is due to the arrangement in staggered
a natural non-polarized light. This is clearly rows of the lamellae on both sides of the striae
revealed by the calculated spectra. The order vertical axis, since the effect vanishes on a sym-
R1 is mostly Transverse Magnetic (TM) polar- metrical structure.
ized, whereas the opposite order R1 is Trans-
verse Electric (TE) polarized. As for the spec-
ular order R0 , not only is it little polarized, Mysterious Adonis
but it also shows a very low intensity. The cal- Every family has its lame duck. I would like to
culations and maps of diffracted elds offer complete this presentation by mentioning the
a different vision. Exposed to a polarized TE particular case of the beautiful Morphos adonis.
Polarization Effects 109

(a)
Figure 8.45. SEM image of posterior wing scales of
female Colias crocea. The structure is similar to that
of male posterior wings. Lamellae hardly overlap
and numerous granules of pigment can be found
between striae, like in the male.

and whether from bottom or cover scales, striae


include only one long lamella along the whole
scales. The periods of these striae are different
according to scales. It is 300 nm long for ground
scales and 600 for cover ones, but sections
show a rigorously identical structure. The opti-
cal study of this buttery might have surprises
in store. It is indeed hard to believe that the
bright reection of the wings only result from
(b) dielectric layer thickness. It could rather pro-
ceed from a general effect of the cover scales,
Figure 8.44. Male Colias crocea. (a) Posterior wing which sometimes overlap three or four times.
scales show the multilayer structure of striae. (b)
Anterior wing scales with lamellae presenting a re-
However, no interferential effects can occur
stricted overlap. In both cases, one can note the pres- anymore since the various thicknesses are sep-
ence of pterinosomes in inter-stria spacings, which arated.
generate the orange-yellow color.
Iridescence in Ultraviolet
It looks quite similar to M. sulkowsky or M. Iridescent effects in ultraviolet are widespread
zephiritis. And yet, its scales structure is com- among Lepidoptera, especially in the Pieri-
pletely different and the origin of its brightness dae big family, the colors of which commonly
remains unclear. derive from pigments or scattering. We are
As has been seen on Figures 8.35 and 8.36, here approaching the limits of human vision,
cover scales are very large, covering the whole thus of colors. Yet, as has been seen, the sen-
wing so that one must remove them to re- sitivity spectrum of butteries extends on a
veal ground scales. They vanish totally in an large part of ultraviolet, an advantage they
index liquid and dont include any pigment. use. The structures responsible for these effects
Ground scales on the contrary are smaller and are surprisingly close to that of Morphos and
highly pigmentary. Nothing extraordinary so should therefore be classied as 2-dimensional
far, yet it is the striae structure that is remark- structures. They are here always associated
able. Contrary to what has just been described, with pigments and scattering structures in the
110 Chapter 8 2-Dimensional Structures: Interferences and Diffraction

Figure 8.46. A few examples of butteries showing iridescence in ultraviolet. They are all widespread in
France, excepting Phoebis rurina from Brazil.

visible, which makes these butteries particu- rior, posterior, or even both. On the other hand,
larly interesting, yet complex. To be able to dis- it insures a certain distinction between mimetic
play characteristics that predators cannot see species, which could otherwise be taken in by
presents at least two advantages. On the one each other, hence hindering reproduction.
hand, it results in a better sexual recognition This is true concerning both Batesian and
between partners of a same species showing Mullerian mimicry and can affect males as well
the same colors and motifs, whether cryptic as females.
or mimetic, in the visible. The differences be- Spectrometric measurements performed on
tween males and females are generally found the anterior and posterior wings of female
on large areas of wings or whole wings, ante- and male Colias crocea underline the different
Polarization Effects 111

cording to the angle of incidence, which re-


minds us of Morphos and suggests that the ori-
gin must be interferential. Depending on the
method used, the adjustment of the shift ac-
cording to Bragg law leads to 105 15 nm
layer thicknesses inclined of 20 on the wing
plane.
This is strengthened by scanning electron
microscopy. Among two species, the male iri-
descent wings contain Morpho-type structural
cover scales, the striae of which consist of
8 to 9 lamellae (L/1 = 8 10) overlapping one
Figure 8.47. Reection factor. Shift of the reection another and substantially tilted ( 20 ) and
maxima and minima of Gonepterix cleopatra anterior remarkably regular, especially in Colias. Their
wings with the angle of incident. The pigmentary average thickness is around 90 nm. On ante-
reection reaches a plateau in the visible, the ampli- rior wings, lamellae are much shorter (L/1
tude of which slightly varies, but not its position.
2), and almost dont overlap. In both cases, the
cavities formed between striae include pteri-
optical properties in ultraviolet. The male pos- nosomes, ovoid small vesicles of pigment that
terior wings show a broad reection peak generate visible coloration.
centered on 350 nm in ultraviolet and are Among females, the structures of anterior
slightly more reecting than female ones in and posterior wings scales, whether ground or
the visible. This peak is neither encountered on cover, is similar to males Gomepteryx cleopatra
male anterior wings, nor on any female wings. anterior ones and Colias posterior ones. They
Gomepteryx cleopatra is the opposite, the male all present a very low ratio of the covering of
anterior wings being reecting in ultraviolet. striae and numerous granules of pigment in
The origin of the motifs in ultraviolet can inter-striae cavities.
be diverse. We will see in Chapter 10 that the The species possessing visible motifs in ul-
common blue Polyommatus icarus shows a scat- traviolet must be widespread, yet their study is
tering peak centered on 380 nm, which marks incomplete. Most of them belong to the Pieride
the limit of our perception spectrum and ex- family and some to Lycenide. Flowers, which
tends well into ultraviolet. It thus includes depend on insects to reproduce, also use visible
motifs from this spectral eld. But, one can motifs in ultraviolet to distinguish themselves
note that the peak in Colias crocea shifts ac- and attract pollinators.
9
3-Dimensional Structures: Crystalline
Diffraction

Theory Recalls a slightly different form, known as Bragg re-


lation and which is based on the complemen-
The structure now presents a periodicity in the tary of the angle of incidence, that is here angle
three spatial directions, and it is under this form  between the incident ray and the reticular
that the analogy with classical photonic crys- plane:
tals is the most obvious. Such photonic crys-
2 d sin = k (9-1)
tals are naturally present in the mineral world
(they are opals), and also in the animal world This approach is simple and speaks for it-
among some butteries, but especially among self. And yet, if it succeeds in expressing the
certain Coleoptera (Curculionidae and Eupholus, directions and colors of diffracted wavelength,
among others). Those Coleoptera possess thick it doesnt mention their intensity, which de-
resisting scales3 to 3,5 mthat are irregu- pends on the nature and arrangement of atoms
larly arranged on the elytrons, limbs, and head. in the planes. That is why we will briey de-
When individually examined with optical mi- scribe, without elaborating on a demonstra-
croscopy, these scales present an extraordinary tion, a more modern and comprehensive ap-
opalescence, producing bright green or blue proach.
colors. A section or some broken scales show a Whatever the object observedthin layer,
very regular 3-dimensional alveolar structure grating, or crystalthe physical phenomenon
with an approximately 200 nm-long period. As is fundamentally the same. An electromag-
a crystal diffracts X-rays, this structure diffracts netic wavelength falling on a point object is
certain visible wavelength in different spatial diffracted in all directions. Depending on the
directions, resulting in the iridescence of the way the punctual secondary sources are related
insect. spatially, these diffracted wavelengths interact
We have followed a historical distinction differently and can either be reinforced or de-
between interference and diffraction so far. stroyed, thus causing interferences in the rst
Diffracting 3-dimensional gratings or crys- case or diffraction in the other. Nevertheless,
talline diffraction perfectly illustrate the funda- the common origin of all these phenomena ob-
mental equivalence between the two phenom- viously lies in the initial emission of a point
ena. A simple way of addressing the problem source. So, the rst thing to specify is the am-
has been to assimilate the crystal to its retic- plitude emitted by the source in any point in
ular planes and to treat the wave diffraction space. The problem can be simplied by mov-
by atoms as the interferences of the reected ing away from the source, which is justied
wave in all directions by these planes. Calcu- when one studies buttery colors produced by
lations are quite similar to those mentioned in molecules or by structures of size around sev-
Chapter 7 concerning interferences produced eral tens of nanometers.
by thin lms, so we wont repeat them here. One shows that at innity, or in the image
However, for historical reasons, the result is in focal plane of a lens, the amplitude emitted by

112
Theory Recalls 113

Figure 9.2. Diffraction at innityor to a lens focus.


The diffraction phenomenon observed in the image
plane is obtained by the Fourier Transform of the
distribution of amplitudes in the object plane.

Figure 9.1. Bragg model. If the incident wavelength


is smaller than or equal to the grating period, an a formalism when studying crystalline diffrac-
interferential approach allows to determine the color tion lies in several simple properties of Fourier
and direction of the diffracted waves. transforms. In particular, that the Fourier trans-
form of a linear combination of functions is the
a punctual source M is the Fourier transform linear combination of the Fourier transforms of
of a functionif the source is situated in the each function. In other words, the calculation
object focal plane. of the diffraction by a grating, whatever its di-
 mension, can be reduced to that of the diffrac-
f(u, v) = eiK(u + v) d d (9-2) tion of its basic motif; the latter function has just
T then to be convoluted with one or several peri-
We wont elaborate on the study of Fourier odic functionsknown as combsindicating
transforms. Undergraduates are familiar with repetition in the motif space.
it, yet it wont much affect people who arent! To nish with this introduction, let us show
Let us just say that the only advantage of such now the calculation of the diffraction gure of a

(a) (b)

Figure 9.3. Polarization effects observed on Curculionidae scales, under natural light (a), and polarized light
(b).
114 Chapter 9 3-Dimensional Structures: Crystalline Diffraction

Figure 9.4. The South African Curculionidae Cyphus


hancocki 2cm. The insect colors are not very bright,
contrary to its isolated scales.
(a)

complex motif crystal reduces to the following


operations:
Fourier transform of a single atom

Fourier transform of the delta functions in-


dicating the atom positions in the molecule

Fourier transform of the delta functions in-


dicating the positions of the molecule in the
crystal.

(b)
Polarization Effects
The rigorous treatment of the polarization of
wavelengths refracted by a grating or a crystal

(c)

Figure 9.6. Organization and structure of Curculion-


idae Cyphus hancocki scales. Like in butteries, one
can distinguish two types of scales that unevenly
Figure 9.5. Photonic microscope image of opales- cover the elytron surface (a). A broken scale exposes
cent colors in Cyphus hancocki. The various colors the internal structure. One can very clearly distin-
of a scale correspond to areas that are oriented in guish reticular planes (b). The grating is tetraedri-
different directions within the structure. catetrahedral (c).
Insects and Photonic Crystals 115

(a) (b)

Figure 9.7. A natural Australian opal and its crystalline structure, left. It is less ordered than in Curculionidae
scales. Photo and collection E. Fritsch and A. Barreau, Institut des Materiaux, Universite de Nantes.

is generally intricate, and wont be tackled here. polarized than others and can be suppressed by
One can convey an idea of it, in terms of qual- using a polarizer. Chromatic effects are some-
ity at least, by keeping a Bragg interferential times quite impressive locally. And yet it is
approach and reminding one that on a diopter, more often difcult to distinguish them at a
the s and p components of a wavelength are macroscopic scale, since natural light is little
reected differently. In a similar way, when a polarized and the scales orientations are quite
crystal is illuminated more or less diffusely, the anarchical.
color observed in a given direction results from
the colors created through interferences of re-
ected wavelengths under various incidences, Insects and Photonic Crystals
and thus differently polarized. Some chromatic
componentsas reected under an incidence The formalism that Bragg developed in or-
close to that of Brewster angleare thus more der to analyze the diffraction of X-rays by a
crystal dates back to the early 20th century,
whereas passion for photonic crystals is a very
recent phenomenon. One should be wary of
too-simple transpositions and make sure that
basic hypotheses remain unchanged. Here, the
initial hypothesis is that structures are actually
periodic on rather long distances and that the
period of the grating and the incident wave-
length present approximately the same size. As
the following illustrations will show, these con-
ditions are met at the scale of the insect scales,
but not at the scale of the insect itself. Although
colors are beautiful locally, the insects tend to
look quite dull.
A good illustration of this is provided by
certain members of the Curculionidae big fam-
Figure 9.8. Two Curculionidae. Eupholus sulcicollis
from New Guinea, on the left, and Eupholus humer- ily, as represented by weevils in our regions.
alis from Papua, right. Like the rest of the family, They can actually show quite nice blue or
they are entirely covered with scales. green colors in other regions. The whole body,
116 Chapter 9 3-Dimensional Structures: Crystalline Diffraction

including legs, head, antenna . . . , is covered by undulations and they dont take part in the
thick and hard scales that seem to be scattered coloration of the insect. When they are me-
in an anarchical way. Two types of scales can chanically broken after freezing, they reveal a
be distinguished: on the one hand, rather at quasi-perfect tetrahedral crystalline structure,
scales plating the elytrons surface and on the actually very close to that of opals, and often
other hand, rare ones that are pilliform and leading to the same iridescent effects. Though
bristled. The former present various shapes- rarer, similar structures were found among cer-
spherical, more or less rectangular or oblong tain butteries such as Parides seostris by Pete
and can also include an indistinct surface Vukusic and Roy Samble from Exeter Univer-
structure. They seem to be striae rather than sity in England.
10
Amorphous Structures: Scattering

Theoretical Recalls Single Scattering, Multiple Scattering


For very weak particle concentrations, i.e., if
Scattering by Particles
particles are small and very far from one an-
The scattering of an electromagnetic wave by other, one can rather accurately consider that
particles is a major source of color in nature. each of them is illuminated by the only in-
The color of eyes, sky, snow, of certain feathers, cident ux. Indeed, a particle scatters a cer-
and of course, numerous buttery wings show tain amount of the incident light in all direc-
that the phenomenon can occur on hundreds of tions. The further one moves away from the
kilometers as well as only a few micrometers. scattering particle, the weaker the inuence of
Obviously, there are various regimes of scat- scattering, as it is spread over a larger sur-
tering depending on the size and concentration face. Thus, beyond a certain distance , the
of the scattering particles. Many refer to famous scattered wave can be neglected, and a sec-
physicians. Very small particles, little concen- ond particle situated beyond wont be af-
trated but at a large scale are linked to Raleigh fected by the former. The problem is in this
senior, who has already been mentioned, way substantially simplied, since one has
whereas important concentration at a very only to multiply the calculation performed
small scale relates to Thyndall. The wide range for each of them by the number of particles
of characteristic lengths of the phenomenon and thus nd the optical response of the whole
will be further described. At a microscopic scattering medium. This is known as single
level, i.e., at the scale of the scattering parti- scattering.
cle, one must consider the interaction between The probability that a wave scattered by
the incident wave and an object of variable in- a particle doesnt meet a second one de-
dex, size, and shape. At a macroscopic level, the creasing with distance, one can easily un-
static aspect of the problem comes into play, derstand that such a simplication is valid
which entails the number and distribution of provided that the thickness of the scattering
particles, and their possible interactions. medium be not too large compared to the
The rst aspect was rigorously treated by particle spacing. If it is not so, i.e., if the
Gustav Mie in 1908 for a sphere of any size and concentration in particles increases or if the
index, immersed in a nonabsorbing medium. thickness of the medium is too important,
This treatment has been gradually extended the scattering becomes multiple. One will al-
to other geometries. It allows us to underline ways be able to treat a single particle, yet
the various parameters involved in the phe- the ux incident on it becomes more com-
nomenon. The second one requires more calcu- plex; it proceeds from both the main source
lations, is still being developed, and allows us and the other particles behaving as secondary
to distinguish the different scattering regimes. sources.

117
118 Chapter 10 Amorphous Structures: Scattering

Figure 10.3. Dependent scattering: particles are next


to each other and their individual response is no
longer that of a sphere.

Dependent ScatteringI, Independent


Scattering
In the cases previously mentioned, scattering
being either single or multiple, the individual
properties of each inclusion were implicitly de-
Figure 10.1. The green color of this grasshopper is termined by their shape and index, indepen-
a complex color due to scattering and to absorption dently from the presence of neighboring inclu-
by pigments. sions. The only incident ux on the particle

(a) (b)

Figure 10.2. Single scattering (a). Particles are far away from one another and the thickness of the scattering
medium is weak; each particle receives only the main incident beam. Multiple scattering (b). The particle
concentration is higher. Their response is still that of a sphere but they are also subjected to the eld scattered
by neighboring inclusions.
Theoretical Recalls 119

could in these cases be affected and undergo


variations. This case is known as indepen-
dent scattering. Such an approximation is ob-
viously not always applicable. The borderline
case in which inclusions touch each other is
enough to demonstrate this. The aggregate
thus formed scatters as an individual entity
that is all but spherical. The radiative charac-
teristics of each particle are modied by the
presence of neighbors that are too close. This is
known as dependent scattering, the latter oc-
curring mainly when particle concentrations Figure 10.4. The basic concepts of the Mie The-
are high, independently from the size of the ory. The spherical inclusion with radius can present
scattering medium. a complex index, but it is immersed in a non-
The physical process and the treatments used absorbing medium with real index n1 . The relevant
parameters are the ratio of n/n1 and the Mie parame-
for modeling such media are very different ter x = 2n1 r/. It allows us to calculate the complex
from one regime to the other, mostly by taking amplitudes scattered in a : S() direction.
into account the phase of the scattered waves. If
congurations are such that interference phe-
nomena can occur between two neighboring the indicator of scattering. The fundamental
inclusions, i.e., if distances between particles parameter of Mies theory is the size param-
are approximately equal to the wavelength, or eter x linking the incident wavelength in a
smaller, one must take into account the phase medium of index n1 to the radius r of the parti-
difference between the various waves and ap- cle:
ply a coherent treatment. If not, a more basic
x = 2rn1 /. (10-1)
calculation, dened as incoherent, involving
only the eld intensities, represents a good ap- Calculations thus allow us to determine the
proximation. electromagnetic elds in every point in space,
i.e., the elds absorbed by the particle as well
as those scattered outside. One can also deduce
Scattering by a Particle
from these calculations energy quantities cross-
Mies Treatment sections, which are better indicators of the scat-
Whatever the scenario, one is brought to con- tering amplitude than the very elds. This is
sider the interaction of a wave with a particle how one denes the absorption cross-section
at the microscopic level. This problem has been
treated by Lorentz as early as 1890, and rigor-
ously formalized by Gustav Mie in 1908 for a
spherical inclusion of any size and index, im-
mersed in a nonabsorbing medium with real
index. We wont introduce Mies calculations
here, which are more complex in their form
than in their content because of their obvi-
ous dissymmetry; the incident wave is plane
and the inclusion is spherical. As for study-
ing the interaction, one must introduce one
forcefully!in the location of the other, i.e.,
here, to split the plane wave into an innite
series of spherical waves. All the rest is tech- Figure 10.5. The human eye iris is another example
nical . . . and allows us to determine the ampli- of a complex color proceeding from absorptionby
tude of the scattered elds in all directions, i.e., melaninand scattering.
120 Chapter 10 Amorphous Structures: Scattering

Figure 10.7. Selective scattering. Particles with


radius approximately equal to the blue wave-
length tend to scatter the latter to a large extent,
while hardly scattering long wavelengthsor red
Figure 10.6. Scattering efciency of a non-absorbing ones. The transmitted light is predominantly red,
sphere in relation to the size parameter x = 2nr/. whereas light that is scattered in all directions is
Scattering reaches its maximum for x 6, i.e., predominantly blue.
for the radius of a particle immersed in vacuum
approximately equal to the wavelength.

or absorption efciency Q, which is the quan-


Cabs and scattering cross-section Csca as the ra- tity representing the phenomenon the best (Q =
tios of the electromagnetic energies absorbed C/a2 in the case of spherical particles).
and scattered by the sphere and the energy of As is often the case when using spheri-
the incident wave respectively: cal coordinates, the expressions of these ef-
 ciencies are formally complex and to present
Csca = Wsca /Iinc
. (10-2) them would be pointless. It is nevertheless pos-
Cabs = Wabs /Iinc
sible, if x is small i.e., when inclusions are
These cross-sections directly depend on the small compared to the wavelength, to expand
size parameter x and thus allow us to evaluate these expressions as power series of x. This is
the relative inuence on scattering of the inci- much more relevant and allows us to demon-
dent wavelength, of the particle size and index strate how scattering can generate colors. If the
within the various regimes studied. Divided product |m| x, where m is the relative index
by the geometric cross-section, these cross- of the particle in its surrounding medium, is
sections allow us to at last dene the scattering small compared to 1, only the rst terms of each

Figure 10.8. It is hard to obtain red


through scattering because of the
dissymmetry of the scattering ef-
ciency. When the radius of parti-
cles is close to blue wavelengths
size parameter : x = 2 6-(a) blue
is highly scattered, as well as UV,
which doesnt affect color. (a) The
other colors are very little scattered.
If the radius of particles is close to red
wavelengths-the size parameter re-
maining close to 6-red is highly scat-
tered but the other colors also ex-
ist, and the color therefore tends to
white (b).
Theoretical Recalls 121

tion of the incident wave, while the shortest


ones (blue dominant) are more present in the
direction of the scattered light. When still ne-
glecting the spectral variations of indices, with
constant scattering efciency, an increase in
the size of particles leads to an increase in the
scattered wavelengths; the scattered light thus
tends towards white.
When the size of the particles increases,
most of the largest wavelengths are succes-
Figure 10.9. Scattering by particles of size close to sively scattered, yet the resulting colors are
red wavelengths. The latter are highly scattered, and never as bright as in the blue. This is due to
so are the others, but to a lesser extent. The aspect is the dissymmetry of the scattering cross-section,
whitish in the forwardtransmission, as well as in which quickly tends towards 0 when the wave-
the scattered directions.
length increases (or when the particle radius
decreases), but tends towards a nite value
of these developments are signicant and are
(and not insignicant: 2) when the wavelength
equal to:
decreases. If one considers particles with a ra-
  dius equal to that of the wavelength (x 6,

8 4  1 2  close to the cross-section maximum), only the
Qsca = 3 x  + 2 

blue wavelengths are scattered by small parti-
1 2
  , (10-3)

1 2 cles in the visible. Ultraviolet is also scattered

Qabs = 4xIm but it doesnt take part in color. In the reverse
1 + 22
case of bigger particles, of size comparable to
where 1 and 2 are respectively the dielectric the red wavelengths, the latter are highly scat-
functions of the particle and of the surround- tered. And yet, so are all the othersto a lesser
ing medium. In this way, if these dielectric extentand the resulting color tends towards
functions dont vary very much in the stud- a slightly colored white.
ied spectral range, the scattering and absorp-
tion efciencies vary respectively like 1/4 and
1/. We here come across again with Raleighs
theory, which shows that the blue wavelengths
on one end of the spectrum ( 400 nm) are
16 times more scattered than reds on the other
end ( = 800 nm). The largest wavelengths (red
dominant) are thus encountered in the direc-

Figure 10.11. The scattering plane is determined


Figure 10.10. Scattering by pigments. In the trans- by the directions of incidence and observation.
mitted and scattered spectra, pigments absorb some It is equivalent to the plane of incidence for the
wavelengths, thus changing color to a large extent. reection on a diopter.
122 Chapter 10 Amorphous Structures: Scattering

Figure 10.12. The feathers of the parakeet Melopsi-


tachus ondulatus are blue due to scattering and yel-
low due to the presence of pigments. There are green
areas when the two sources of color overlap. Figure 10.13. Scattering maps of waves polarized
parallel (----) and perpendicular (----) to the scat-
Scattering with Pigment tering plane, and the result for a nonpolarized wave.

In nature, scattering particles are rarely de- wavelength, they are as follows:

prived of pigments. A widespread case is 9 |a1 |2
i// = cos ,
2

melanin, which often appear as granules, like 2
4k r 2 1 
i= i// + i .
in the iris in the eye, for instance. As will 9 |a1 |2
2
be shown, melanin is a pigment ranging i = ,
2
4k r 2
from black-brown to yellow, i.e., absorbing the (10-4)
smallest wavelengths, and mostly the blue. If where k is the wave vector, r and are the
the size of granules is such that it produces polar coordinates of the measurement point
scattering, the scattered spectrum as well as in a location, the center of which is occupied
the transmitted one are here deprived of any by the scattering particle, and a1 is a scatter-
blue wavelength. Usually containing a lot of ing coefcient depending on the size parameter
blue wavelengths, the former thus turns more and the particle index. Let us only note with-
green. The second one, which was already de- out elaborating that when the incident wave
prived of blue wavelengths, is not altered much is polarized perpendicularly to the scattering
and remains red. This combination of effects plane, scattering is isotropic, but that it varies
gives birth to many green colors, such as that like the square of the cosine of the scatter-
of lizards, grasshoppers, and many others. ing angle when polarized perpendicular. This

Polarization Effects
Polarization effects generated by these scatter-
ing phenomena are important in many regards.
They can be found in both the light scattered by
insects, which gives them their colors, and also
light that sometimes illuminates themsunlight,
for instance. The directions of polarization of a
scattered wave are marked within the scatter-
ing plane formed by the direction of the inci-
dent ray and that of observation.
My calculations that have only been briey
mentioned allow one to determine the scat-
tered intensities in a given direction for a polar- Figure 10.14. The common Pieridae, Pieris brasicae,
ization state of the incident wave. For a given female.
Type Butteries with Scattering Structures 123

implies that even when the incident wave is not scattering that can both substantially split col-
polarized, part of the scattered wave is polar- ors, scattering is relatively little selective. This
ized and always perpendicularly to the scatter- is easily understood when one remembers that
ing plane. The P polarization ratio, dened as a particle of a given size will highly diffuse the
the following product: radiations of wavelengths that are shorter to
i i// itself, whereas it wont diffuse much that of
P= , (10-5) longer wavelengths. Consequently, scattering
i + i//
by itself cannot produce a color situated at the
becomes, according to the previous equations: center of the visible spectrum or at its red end.
Very small particles can indeed diffuse blue
1 cos2
P= , (10-6) wavelengths but hardly others. Nevertheless,
1 + cos2 when one increases their sizes, they wont dif-
which demonstrates that at right angle of the fuse just a greater wavelength, but more and
incident direction, the scattered wave is even more wavelengths, from green to yellow . . . the
totally polarized. This effect is similar to the superposition of which leads to an increas-
total polarization obtained in reection at the ingly pure white. This process implies a mixing
Brewster angle. The scattering diagrams of rather than selection of colors, and the resulting
waves polarized perpendicularly and parallel colors are not very pure. More marked colors
to the scattering plane are shown on Figure will appear, provided scattering be combined
10.13, as well as their result when the incident with a more selective phenomenon, which is
wave is not polarized. pigmentary absorption.
We will illustrate this phenomenon with two
Type Butteries with Scattering species that live in our temperate regions and
that are situated at the two ends of the spec-
Structures
trum, as has just been mentioned: the male
Common blue Polyommatus iracus and Celas-
Pieridae and Lycenidae
trina argiolus on the one hand and the white
White or Blue Pieridae Pieris brasicae on the other hand. As
Scattering as a source of colors is widespread has been already said, those are the only colors
in nature. Although it may seem quite labo- that can be obtained through scattering. But the
rious, the presentation we have just made of two species also happen to illustrate the two
the phenomenon enables to convey its intri- possible ways of scattering: a high index par-
cacy. After this long chapter on thin layer in- ticle immersed in air or, on the contrary, an air
terferences and the fact that theoretical pre- inclusion in a material with a higher index.
dictions and experimental conclusions corre-
spond, the present chapter may appear quite
short and disappointing. Yet, it will be justied
by the following study of the wingaccording
to the same zooming in approach. Whether it is
from the microscopic point of view, from that
of each scattering particle, or the macroscopic
and static point of view of their distribution
on wings and their interactions, formalism is
much more complex than what has been seen
so far. As cross-sections observed by electron
microscopy will show so wonderfully, one is
here confronted with an extraordinary accu-
mulation of mathematical and calculation com- Figure 10.15. Diffuse reection factor under normal
incidence of a Pieris brasicae anterior wing. Reec-
plexities, these butteries thus proving impos- tion is strong throughout the visible spectrum.
sible to model so far. Let us come back from UV and part of the blues are absorbed, hence the
the beginning. Contrary to interferences and slightly yellow color of the wing.
124 Chapter 10 Amorphous Structures: Scattering

(a)
(b)

(c)

Figure 10.16. (a) SEM image of the extremity of a Pieris brasicae scattering scale. (b) view of the center of the
scale approximately showing a non-covered area and the neighboring scales. In covered areas, striae are
denser, counter-striae are thicker, and inclusions are fewer. Pieride scales also present traces of leucopterine,
mostly on the ventral side of posterior wings, which gives them a slightly yellow color. Pterinosomes: (c).

Figure 10.17. Scale roots covered by


the extremity of preceding scales.
Roots contain fewer Pterinosomes.
Type Butteries with Scattering Structures 125

whether in reection or in transmission and


showing quite similar shapes. The whole ap-
pears rather disorganized and scales largely
overlap.
By using scanning electron microscopy,
striae look little developed, with an average
spacing of 1,5 m. Lamellae are very short,
do not overlap, and therefore cannot pro-
duce any interference effect. They include re-
markably developed counter-striae, which are
regularly spacedapproximately 600nmthe
whole creating a surprising square pattern on
the wing formed by parallel rows of rectan-
gular compartments. The latter contain scat-
tering particles, around 20 per compartment.
The particles are tiny pterin granules, known
as pterisosomes, and present a regular and
slightly ellipsoidal shape, all oriented vertically
over the surface of the wing. The small axis is
100 to 150 nm long and the largest isaround
350 nm, although it is quite difcult to estimate
since its base is not always visible.
In this case, we are confronted with ellip-
Figure 10.18. The male blue Argus Polyommatus soidal scatterers that are arranged on a plane,
icarus and the Argus Celastrina argiolus. The colors of parallel to each other and almost touching.
both mainly proceed from scattering of the shortest Given a pterines estimated index of about 1,6
wavelengths. The ventral side of P. icarus is pigmen- in the visible spectrum, the parameter of the
tary, more bluish at the base.
sizes of isolated particles ranges from 3.5 in
blue to 1.6 in red, which doesnt allow us to
White obtain maximal scattering. And yet, particles
The scales of Pieris rapae are long, narrow, touching each other, we are here obviously
indented, and highly inclined on the wing confronted with an extreme case of dependent
plane. It is hard to distinguish between the scattering, which experimentally leads to an
bottom and cover scales, all being evenly white, overall substantial reection on the whole of

Figure 10.19. Diffuse reection


factor of P. icarus and C. argi-
olus Argus. The former shows
a strong reection factor cen-
tered on the blue/ultraviolet
limit. While the second, present-
ing bigger inclusions, scatters
more in the visible and has a
weak spectral purity.
126 Chapter 10 Amorphous Structures: Scattering

Figure 10.20. Superior side of Poly-


ommatus icarus wing showing at
and highly striated scattering scales,
as well as the root of a long thread-
like scaleover 500-m longand
small androconies in interstice.

the visible spectrum, yet the modeling of which There is no inclusion here. The bottom of com-
remains to be established. partments is lined with an air spongy structure.
It is difcult to evaluate the structure thickness
by using Scanning Electron Microscope (SEM)
Blue images. Yet one can roughly estimate the aver-
The common blue appears like the negative of age size of pores, here below 100 nm. The ef-
Pieris brasicae. Most of the scales, with a more ciency of scattering diminishes very substan-
geometrical shape, prove quite disorganized, tially for long wavelengths, which makes short
while some very long threadlike scales are scat- blue and violet wavelengths become more visi-
tered on the whole. Spacing between striae is ble. The intensity remains weak and the purity
well marked, lamellae are short and not cov- much lower than that of Morphos obtained by
ered, and counter-striae are well developed. interference.

Figure 10.21. Spongy structure of


the scattering layer of Polyommatus
icarus. The structure is highly per-
meable and by using an optical mi-
croscope, one can observe how after
absorption of an index liquid, color
disappears.
11
Pigments and Pigmentary Colors

If the present book is mostly dedicated to phys- energy to the latterand thus disappearif
ical colors, they are still a minority in nature. the energy corresponds to the difference of en-
Most of animal colors derive from a pigment, ergy between two levels of the atom. After a
and although these substances dont produce photon is absorbed, the atom is in an unset-
iridescence or very little in the usual observa- tled stateknown as excitedand goes back
tion conditions, they still have their place in to its fundamental state. Several processes al-
a book dedicated to buttery colors. We will low this. Symmetric to the previous, the rst
mention them along two distinct lines, rst the one, is radiative relaxation. The atom goes back
physicochemical aspect, which will include an to its fundamental state by emitting its excess
overview of some important and distinct fam- energy under the form of a photon. An exte-
ilies of pigments as well as the physical pro- rior observer wont notice anything: A photon
cess causing color: selective absorption. Since enters and another, identical to the rst, goes
buttery colors are observed in reection, this out. The incident light is not modied and no
aspect of the problem, generally less known, color appears. However, such relaxation can be
will be further studied. Secondly, we will tackle non-radiative. Surplus energy can, for instance,
the physiological aspect of pigmentation, as make a crystalline grating or its neighboring
well as its origin and from the point of view molecules vibrate, thus leading to an increase
of evolution. in the temperature of the object. Yet, there is no
emitted light anymore. The incident photon is
Selective Absorption and Colors indeed absorbed during such a process. Yet, for
of Pigments the observer, this or these photons have disap-
peared from the incident light spectrum, and
Absorption is a transfer of energy between the the color of the body somehow corresponds to
incident electromagnetic wave and the atoms the complementary of the absorbed color.
or molecules composing the illuminated object, In order for such a phenomenon of coloration
or more precisely, the electrons of these atoms to occur, the difference of energy between the
that are distributed on various orbitals, each two levels must correspond to a frequency that
characterized by one energy level. is perceivable by the human eye. For us, this
Transitions between the different levels pro- corresponds to wavelengths ranging from 380
ceed from absorption or emission of the corre- to 680 nm. It is generally the case for isolated
sponding energy. In the case that concerns us atoms in a gas, but not atoms that are linked
here, the energy that is absorbed comes from to each other. In this latter case, the absorp-
the incident light, which will thus be deprived tion bands of electrons forming covalent bonds
of certain wavelengths and its color will be con- are usually situated in the ultraviolet. Pigments
sequently changed if the latter are situated in and dyes belong to a very singular category of
the visible part of the spectrum. When a pho- substances, in which the distribution of elec-
ton interacts with an atom, it can transfer its trons is such that absorption bands are brought

127
128 Chapter 11 Pigments and Pigmentary Colors

Figure 11.1. Transition between two energy levels of


an atom with absorption and emission of a photon.

back to the visible. This category is that of or-


bitals. Relaxation on the inferior levels occurs in
a non-radiative manner through heat transfer-
ring, phosphorescence or uorescence, or dur-
ing a chemical reaction with the surrounding Figure 11.3. Optical index of a materialn real part
medium, in which case pigment is decomposed and k imaginary partdue to electronic polariza-
and loses its initial coloration. The function, tion. Other polarization phenomena occurring when
frequency is lower, make the real part dissymmetric
or elementary group, responsible for light ab- on both sides of the absorption peak. The reection
sorption is called chromophore. Its color can factor is thus also different. Absorption in the yellow
be modied by linking with other chemical makes an object look red in reection.

groups receiving or giving electrons; they are


auxochromes.
Organic chromophores always present a se-
ries of carbon atoms that are alternatively
linked by single or double bonds. The sim-
plest example of this is the benzene cycle. The
two representations with localized bonds are of
course strictly identical and equiprobable, and
the electron pairs forming them should only ab-
sorb ultraviolet radiations. However, none of
these representations is adequate. The most re-
alistic one consists in linking atoms with single
bonds, the supplementary three atomic pairs
being divided and distributed on the whole
molecule in delocalized orbitals (orbitals ).
Here again, the excited states of these orbitals
correspond to ultraviolet radiations. The en-
ergy of orbitals decreases and their absorp-
tion bands come back into the visible only if
molecules of the type mentioned above gather
and form long polymers. The resulting col-
ors are altered or intensied by the presence
of auxochromes on the polymer margins. The
auxochromes role is to increase the mobility
of chromophore electrons , thus the wave-
Figure 11.2. Different schematizations of benzene
length of absorption. Contrary to that of iso-
cycle. The two (a) and (b) cycles are symmetrical
and equiprobable. (c) is a more appropriate repre- lated atoms, the energy levels of molecular
sentation. Below, the schematized representation of orbitals are relatively high. One could al-
orbitals. most designate them as energy bands, each
Selective Absorption and Colors of Pigments 129

containing many vibrational sub-levels. Their


number varyies with the complexity of the
molecule: the more complex the molecule, the
more the vibrational sub-levels. Absorption
bands are thus also broadened and the colors
obtained are generally not very pure. The en-
ergy diagrams of pigmentary molecules con-
sequently present wide bands, centered in the
visible range of the spectrum or next to it. When (a) (b)
the pigment or the pigmentary solution is il-
luminated by white light, all the wavelengths Figure 11.5. The dihydroxy-5,6 indole monomer
contained in the incident light and correspond- and its three possible insertion points. (a). A portion
of a eumelanin skeleton (b).
ing to a transition energy between two levels
(the fundamental and an excited level in gen-
eral) are absorbed, and the corresponding en- Substances presenting selective absorption
ergy is the more often turned into heat. The such as pigments in the visible also show a
transmitted spectrum is therefore deprived of selective reection phenomenon that is due to
this range of wavelengths and the resulting the asymmetrical refraction index on both sides
color is the complementary of the absorbed of a large absorption band. As has been men-
one. And yet, butteries are hardly observed tioned, unless other processes come into play,
in transmission, but more in reection. In this the refraction index n indeed tends to be higher
case, the reected color is not the complemen- on the red side of the band, thus producing
tary of the absorbed one anymore: This is se- a higher reection, compared to the blue side
lective reection, a phenomenon so dear to (Figure 11.3).
Michelson, which will now be described. The colors created due to their dominant
reection factor on one side of a pigment
absorption band, also depend on the angle of
incident. Yet, this phenomenon can only be
clearly perceived under polarized light parallel
to the plane of incident. In this case indeed, the
Brewster angle of long wavelengths radiations,
an angle for which the reection factor goes
through a minimum, is approximately higher
than that of blue radiations. Thus, the latter,
mostly responsible for color under normal
incidence, are here almost not reected under
high incidence (from around 60 to 70 ), which
cause a shift of the observed color towards
blue, i.e., an effect similar to that of an inter-
ferential layer. Hence the controversy between
Michelson and Raleigh. This phenomenon
is of course undermined in non or weakly
Figure 11.4. Variations of the reection factor with
the angle of incidence, for two radiations situated on polarized light and cannot totally explain the
both sides of an absorption peak. The optical index important iridescent effects as observed in
is weaker in the blue than in the red of the spectrum, butteries in natural light.
and so is the B pseudo Brewster angle. Under a light
polarized parallel to the incident plane, the reection
factor of the blue wave Rp is weak for incidences Lepidoptera Pigments
close to Bb and the object looks red. The reverse
occurs when incidence is higher and the object turns Among diurnal butteries, excluding melanins
more blue. that form the alary pattern motifs and that are
130 Chapter 11 Pigments and Pigmentary Colors

Figure 11.6. The pheomelanine monomer and its ex-


tension points.

present in all species, each family character-


izes itself by a given type of pigment. They
are mostly ommochromes among Nymphalides,
papillochromes among Papilionides, and pter-
ines among Pierides, while avonodes and bil-
iary pigments can be found in various species.

Melanins
Melanins form a large family of miscellaneous
pigments, widespread in both vegetal and
animal realms. They certainly represent the
most common, yet most complex, pigments. Figure 11.8. Albinism occurs when melanin is not
synthesized.
They are polymers, the biological precursor Here, Pieris brasicae. Above, the fellow individual
of which is tyrosine. Their colors range from is normally pigmented, white areas are colored by
yellow to brown and black. In the animal pterines and the apex by melanins. The individual il-
world, there are two distinct kinds: phaeome- lustrated below, presents normally synthesized pter-
lanins (yellow to reddish) and eumelanins ines, but the apex scales are depigmented due to the
incapacity of synthesizing pterines. (11.7 and 11.8
Photo and collection H. Descimon)

Figure 11.7. Melanism examples among Melanargia


galathea.
Above, leftnormal individual.
Above, rightsuffusion of black scales in the white
Below, leftmelanian individual extension of black Figure 11.9. The absorption spectra of the
motifs. monomers of the two animal melanins. They
Below, rightgeneralized melanism; black motifs usually cover ultraviolet to a large extent, reaching
still remain visible. the green-yellow range in the visible.
Selective Absorption and Colors of Pigments 131

(brown to black), while Allomelanins are only Pterines


found in Plants, Mushrooms and some Bacte-
Pterines are well known components. One pter-
ria. Phaeomelanins and eumelanins are often
ine is contained in a major vitamin, folic acid,
associated and generate a great diversity of pig-
and others are coenzymes. Their biosynthe-
ments. All the reactions causing melanins are
sis chain is also well known, mostly thanks
catalyzed by a unique enzyme: phenoloxydase.
to experiments performed on butteries (Pieris
Yet the latter are very sensitive to environmen-
and Colias). They proceed from the molecular
tal conditions like temperature, humidity, or
redistribution of nucleotide, the triphosphate
CO2 rate. Variations in these parameters can
guanosine (TGP)the TGP purine nucleus and
lead to cases of melanism or albinism, as has
the ribose merge into a double heterocyclic nu-
been mentioned, directly affecting the survival
cleus; that, is pteridine. Secondary enzymatic
of the species or some of its members.
reactions generate pterines with varying lat-
A key element of human pigmentation, opti-
eral chains. Among them, one can distinguish
cal properties of eumelanin and phaeomelanin
leucopterine and isoxantopterine, which are
solutions are well known. However, the struc-
white, xanthopterine and sepiapterine (yel-
ture of the polymer has not been precisely es-
low), and erythropterine (orangey red). The
tablished yet. The extinction coefcient of the
deep red of certain Appias nero is due to a
monomer always shows a maximum in ultra-
pterine dimer, that is pterorhodine. All of these
violet and then rapidly decreases in the vis-
pigments almost only concern Pieridae and they
ible, hence the yellow-brown colors that are
are situated in scales as microscopic granules,
generally observed. Yet it is established that
called pterinosomes.
the lengthening of the polymer chain and the
presence of auxochromis increase absorption in
the visible, sometimes leading to deep black. Ommochromes
In addition to their role in the alary pattern, Yellow, red, orangey, tawny, or brown. They
melanins often serve as an opaque screen un- are quite complex molecules deriving from the
der or within structural scales. As has already polymerization of cyclic amino acidsmostly
been seen, buttery wings often contain sev- tryptophanethat are found in the eyes of a
eral layers of scales of different kinds and the majority of insects and on the scales of numer-
alary membrane itself is not optically inactive. ous Nymphalides. One owes its study to the Ger-
In the cases of colors generated by the dorsal man A. Butenandt and his group. As they were
side scale pigments, for instance, and observed mostly chemists, their work lacks a zoological
in reection, the transmitted light can be par- study of the distribution of these components
tially reected by the alary membrane or ven- among butteries. In this way, the determina-
tral side scales. It thus once again goes through tion offered here, based on elementary criteria,
pigmentary layers almost without any color al- remains to be proven.
teration, since it has already been deprived of The simplest components are soluble in wa-
the absorbed colors, and thus contributes to the ter. It is the case of the initial component of
globally reected light by weakening its purity. ommochromics biosynthesis chain; that is, the
This phenomenon surprisingly scarcely occurs yellow uorescent 3-hydroxycynurenine. One
with structural colors. With a few exceptions can nd it on wings of some Helioconius. The
like Morpho godarti and Morpho polyphenus, the same butteries present xanthommatine that is
affected butteries always present a layer of orangey and the red dihydroxanthommatine.
highly absorbing pigmentary scales (Archeopre- Vanesses show the latter, but also D ommatine
pona type) or structural scales containing pig- (red-brown) and rhodommatine (dark red).
ments themselves (Morpho menelaus type) and The only difference between these components
creating a dark screen, hindering any parasitic is their lateral chain. Yet, the origin of the bright
reection that would undermine the purity of red pigment of Melitaea didyma and of cer-
the reected color. tain Cymothoe or Callicore remains a mystery.
132 Chapter 11 Pigments and Pigmentary Colors

Leucopterine

Isoxanthopterine

Xanthopterine

3
Sepiapterine

Erythropterine

(a) (b)

Figure 11.10. Pterinosomes between the striae of a Hebonia glocippe Pieridae.

Ommines are little soluble in water and it is dopa) for the simplest ones. The structures
quite difcult to study them from a chemi- of other more complex ones are little known
cal point of view. They are mostly found in and only the structure of papiliochromic II has
eyes. Other pigments, red and highly insoluble, been discovered. Their coloration ranges from
that can be observed in Nymphalides (Precis light yellow to red, and as their names suggest,
coenia), prove to belong to the same chemical they are present in Papilionides. A polymor-
family since they derive from tryptophane. phism determined by one gene and opposing
It was established through blending it with a light yellow pigment to an orangey one has
tryptophane14 C. been researched among Zerynthia. The Greek
species, Z. polyxena, thus includes an orangey
type, Ochracea, present in both males and fe-
Papiliochromes
males. Among Z. rumina, it is the canteneri type
They are close to ommochromics and de- characteristic of Southern Spain and Maghreb,
rive from cyclic amino-acids, tryptophane which is dominant and tends to only occur in
(via cynurenine) and from phenylalanine (via females.
Selective Absorption and Colors of Pigments 133

(a) (b)

Figure 11.11. Left: two Anthocaris Pieride, A. car-


damines and A. euphenoides, colored by pterines.
White proceeds from a mix between leucopterin
and isoxanthopterine, yellow from xanthopterine
and sepiapterine, orangey to the two previous ones Figure 11.12. Nymphalide from the center of Africa,
mixed with erythropterine. Green is due to the Cymothoe sungarismale, 5 cmone of the very few
pointillist juxtaposition of pterine yellow scales entirely red butteries. Its ventral side is cripteous.
and structural bluish black scales.
pilionidae (Graphiini, Parnassius), and Prieridae
Flavonods Dismorphiinaewhere they are associated with
pterinesNymphalidae Satyrinaelike Mela-
Flavonods or anthoxantines rank among the nargia, which reacts extremely well to the de-
rare vegetal pigments found among adult in- tection technique mentioned above.
sects. They are absorbed by the caterpillar,
then stocked and xated in wings during
Biliary Pigments
nymphal stage. When mixed with ammoniac,
they turn bright yellow, which allows to eas- These pigments are unique as their col-
ily detect them. They are widespread among oration tends to be green or even blue. They
diurnal butteriesHesperiidae, Lycaenidae, Pa- are components deriving from tetrapyrrolic

(a) (b)

Figure 11.13. Flavonoids among various butteries from different genus. Lysandra coridon, Leptidea sinapis,
Melanargia galathea. Above, normal aspect. Below, insects immersed in ammoniac vaporscharacteristic
reaction of avonoids.
134 Chapter 11 Pigments and Pigmentary Colors

Papiliochrome 2

(cf. hemoglobin or cytochrome). Pterobiline


that is usually found in the alary mem-
brane, makes the latter turn green in female
Gonepteryx, in the Cabbage Pieride or even
more in Graphium (papilionidae), and Actias . . .
They are very unsettled pigments and become
altered very rapidly especially in Eurotides
celadon from Cuba, the extraordinary blue of
which turns greenish when it dies . . .
The pigments of diurnal butteries are rather
well known, whereas those of Heterocers are
very little known. They could have many sur-
prises in store if studying them became as com- Figure 11.14. The green spots of this Graphium
mon as it used to be in the past. Even among di- weiskei are due to a biliary pigment, a pterobiline. The
urnal butteries, red pigments remain obscure. light purple also proceeds from a pigment, which
Y. Umebachi and H. Descimon have inde- has still not been identied to my knowledge.
pendently tried to characterize the red pigment
of the Papilionide Atrophaneura horishanus
most probably a papiliochromicyet their The study and characterization of butter-
analysis was hindered by extraction, as was y pigments are far from being complete.
also the case for ommochromic. The same In addition to the fact that they are quite
author isolated the pigment that confers neglected nowadays, these studies are often
the extraordinary bright orange of Coppery confronted with major technique difculties.
among Lycaena phlaeas (Figures 5.1 and 5.2). Indeed, many of these substances are hardly
It is a component deriving from the acid 3- soluble, thus difcult to extract. We will only
hydroxyanthranilique (Umebachi, 1982). And offer an overview here based on a study by
yet, one may have assumed that such a color Pr. Henri Descimon, which includes a compre-
was physical as it is so bright. This discov- hensive bibliography on the subject.
ery is all the more interesting since some The big families of pigments as described
pigments presenting the same chemical na- above come from two different origins. They
ture and colors that are close are common can be the resultor residueboth of a
among mushroomsin Agaric, whether impe- synthesis or an excretion. In most cases, these
rial mushroom or y agaric. components are toxic in high concentration and
must therefore be eliminated or stocked. Para-
doxically, wings, which make butteries look
so beautiful, serve as containers for these tox-
ics. Wings are indeed a dry and sclerotic organ
in adults that can thus stock toxic substances
without any risk. It is also worth noticing that
their coloration is a mere secondary effect, even
if it becomes a key element for the survival of
Pterobiline insects.
12
Thermoregulation and Spectral
Selectivity

Imperatives energy. The wing optical and infrared proper-


ties more or less contribute to each of the pro-
As in the animal world, the body temper- cesses.
atureor that of some parts of the body of Let us rst consider the heating phase. All di-
insectsis one of the principal parameters of urnal butteries have two techniques in differ-
their vital balance. It is indeed directly linked to ent proportions, enabling them to make their
mobility thus affecting feeding, reproduction, thorax muscles reach the critical temperature
and reactions to threats in butteries . . . necessary to get off. The rst process is en-
During ight, it is essentially the six pairs of dothermal, as the muscles themselves maintain
tergosternal muscles, situated in the mesotho- the optimal temperature by stimulating their
rax, that make anterior wings ap, whereas metabolism. This situation is more often ob-
posterior ones are equipped with a mechanic served when the insect is resting in the shade
frenate coupling. These muscles are not di- and is characterized by quick trembling that
rectly attached to wings but permit the thorax resembles human shivering. Inductor and ab-
to contract longitudinally, apping resulting ductor muscles contract quickly and in phase,
only from a mechanic reaction. The efciency i.e., in an antagonist way, thus preventing any
of such a device is surprisingly weak. An es- apping and saving the entire energy to rise
timated 80% (on average) of the energy used the thorax temperature.
during ight is converted into heat. The second process, which concerns us more
The overheating of muscles is not only the directly, is designated as exothermal. It picks
harmful result of muscular activity, but also a up solar energy, thus involving the wing opti-
requirement for ying. During takeoff, wing cal properties but also that of the insect body.
abductor and inductor muscles must quickly The alary membrane proves to conduct heat
and in alternation so they dont collide with poorly and lymphatic circulation is weak, so
each othercontract, the resulting optimal it has been assumed that it is essentially the
temperature reaching 40 C. This heterothermal body or basal areawhich represents less than
animal, which must reach a temperature of a third of the total wing surfacethat picks up
36 to 38 C to be able to take off, is equipped energy. The way wings are situated in relation
with a ying device that is not only ener- to solar ux depends on their optical properties
getically little protable, but that would also and varies according to species. Butteries with
not withstand a temperature over approxi- wings presenting dark colors on either side di-
mately 42 C. However, after taking off, and in rectly pick up solar energy by spreading the lat-
a usual situation, a dynamic balance is found ter to the sun or pressing them vertically above
with a lower temperature ranging from 30 to the body.
35 C. In this way, butteries must be able to The latter position is adopted by butter-
quickly heat their muscles when at rest and ies with cryptic colors, both allowing camou-
then during ight, efciently eliminate caloric age and thermal regulation. White species or

135
136 Chapter 12 Thermoregulation and Spectral Selectivity

Figure 12.2. Solar ux concentration on thorax and


Figure 12.1. Endotherm heating of thorax and ab-
abdomen. In wide open position, only a small part
domen of a Lepidopteron. It takes a little more than a
of energy is concentrated on the body. The smaller
minute for the insect to reach the thorax temperature
the wing angle, the larger the energy concentration.
that is critical for taking off. The slight increase in ab-
dominal temperature during the cooling off phase
suggests that the insect evacuates energy through
trachea from its thorax towards its abdomen. (after or transmitted ux. We get:
B. Heinnick, Sciences 185 (1974), 797)  
Q = Sc Ei Op A0 (Ta T0 ) Sa T4a T40
(12-1)
light-colored ones make their temperature rise where Sa is the collector surface, Sc the sur-
indirectly by concentrating solar ux on their face that is really exposed to radiationthey
dorsal side, which is black in this case. Facing are equal when one considers wings onlyOp
the sun, wings are more or less open above the the total optical loss, Ei incident light, T0 and Ta
body at an angle determining the amount of the temperatures of the environment and of the
energy picked up by the body. collector, and nally and absorptivity and
As concerns cooling that is only required emissivity of the collector.
during ight, most of it proceeds from con- Scales actually prove to be a great insulating
vection and conduction of the environment air, coating, which is why convection and conduc-
and also from tracheal exchanges between the
thorax and the abdomen. Another smaller part
derives from the thermal radiance of the body
and wings. If one considers the average tem-
perature of 40 C, this radiance occurs in near
infrared reaching its climax around 9,5 m. Be-
fore further studying the thermo-optical prop-
erties of the buttery collector, we will now
briey present the few principles involved.

Energetic Balance
A Few Denitions
Figure 12.3. Photothermal collector losses. Optical
Endothermal sources of heat aside, as concerns losses are due to light being reected on wings. They
are very small in black butteries. Losses proceed-
the Q energetic balance of a solar collector, solar
ing from convection and conduction are small when
energy incident on the whole collector surface the insect is at rest, but can be substantial yet use-
represents the credit, whereas the debit is the ful during ight in order to get rid of surplus heat.
convective, radiative or optical loss: reected Losses due to radiation are predominant.
Energetic Balance 137

tion loss for a resting animal can be neglected


(A0 = 0). One can therefore determine a radia-
tive efciency allowing to study the inuence
of the optical properties of buttery wings or
body:
 
T4a T40
r = (12-2)
Ei
One can notice that in a given environment, Figure 12.4. Reection and transmission curves of a
the greater the efciency, the greater the ab- selective collector and an ideal selective window. As
concerns the former, absorption is maximal until the
sorptivity and the weaker the emissivityfor a cut off wavelength c and is equal to zero beyond it.
given body temperature. Let us study these two As concerns the latter, the transmission is equal to
quantities. In a given conguration, the absorp- zero beyond c , preventing thus radiative losses.
tivity of a material is the amount of incident
energy that it absorbs. One distinguishes be-
where L is the monochromatic luminance of
tween the directional absorptivity  () when
the incident uxthe solar spectrum as far as
the radiation specularly falls onto the wing un-
we are concernedand M (T) the monochro-
der an angle (clear sky) and hemispheric
matic emittance of the black body at temper-
absorptivity when the incident radiation is dif-
ature T. These relations tend to be complex,
fuse (cloudy sky), the latter being the integral
but not independent. Indeed, if the body is
over all the irradiating space of the former. As a
at thermal equilibrium, a conservation law
rst approximation, these two magnitudes do
Kirchhoffs second lawdetermines that for a
not depend on temperature.
given direction and wavelength:
As concerns emissivity, it is dened as the
ratio between the energies radiated by the ma-  () =  () (12-5)
terial and by the black body at the same temper-
Absorptivity is directly related to the trans-
ature and in the same conguration. In a given
mittivity  () and the reectivity  () accord-
direction, one measures the directional emis-
ing to the law of energy conservation:
sivity T () which, integrated over the whole
open space, gives the total emissivity T . De-  () +  () +  () = 1 (12-6)
termined in this way, T and represent global
Energetic quantities involved in the radiative
energetic magnitudes in which the absorbed
balance can therefore be obtained through re-
and emitted spectra do not show up precisely.
ection and transmission optical spectrometric
They can nevertheless be deduced from the T ,
 measurements.
() and () monochromatic quantities by in-
tegration on the wavelengths and dividing by
Ideal Selective Collector
the irradiance of the incident spectrum and the
black body emission spectrum, respectively: If the aim is to increase the efciency of the col-
 lector to reach thermodynamic balance, the op-


() = ()L () d

L () d,

timal solution is therefore to realize an absorp-
0  0 tivity maximal over the whole solar spectrum

and an emissivity minimal in the spectral do-
= 1

(, ) cos d,
 main emission corresponding to the tempera-
(12-3) ture reached. For an opaque object ( = 0), this
and comes back to ideally have a reection coef-


 cient equal to zero in the visible and equal to 1 in
T () = T , ()M (T) d

M (T) d, infrared. Equations (12-5) and (12-6) highlight
0 0
that in this case, is maximal in the visible and

= 1  (, ) cos d,
 minimal in infrared. The transition from visi-
(12-4) ble to infrared occurs for a wavelengthcalled
138 Chapter 12 Thermoregulation and Spectral Selectivity

infrared due to the combined action of reec-


tion and transmission. As can be seen in Figure
12.5, the basal areas of the wings turn lighter
black to dark brown when the insect is im-
mersed into an index liquid. This demonstrates
that absorption doesnt depend on pigments
only, but also partly results from a structural
effect.
Solar collectors indeed commonly use the
Figure 12.5. Archaeoprepona menander (Charaxinae) surface roughness in order to increase vis-
iridescent medial areas present an Urania-type scale ible absorption and sometimes even to ob-
arrangement, with convex structural cover scales tain a higher spectral selectivity. The tech-
and multi-lobed pigmentary ground scales. Dipped
nique consists of creating a roughness big
into trichlorethylene, its colors vanish totally and the
black areas of both sides turn dark brown. enough so that short wavelengthsthe entire
visible spectrumget trapped on the ab-
sorber surface, whereas longer wavelengths
cut-off wavelength c depending on the tem- corresponding to the thermal emission spec-
perature reached by the body and which is trum of the collectorare reected.
about 4 m in the present conditions, Tc 50 C. In the industry of solar collectors, this is ob-
This is how most of plane solar collectors tained by producing dendritic structures over
designated as selective, i.e., with marked tran- metal surfaces, providing them with a high
sitions between their visible and infrared op- intrinsic infrared coefcient. The manufactur-
tical propertiesare produced nowadays. Ac- ing process, mainly chemical, only produces
cording to the same principle, we have to men- roughness with random shapes and sizes,
tion transparent coating or selective glass. In which doesnt make them a good support for
this case, visible reection is equal to zero ( = rigorous modeling. They are determined by
0) and transmission to 1 up to c . Beyond, is the statistical distribution of heights in rela-
equal to zero and to 1. Equations (12-5) and tion to the average level and by an autocorre-
(12-6) thus reduce to: lation function transposing the distribution of
  the roughness in the absorber plane: roughness
 () =  () = 1  () +  () (12-7)
spacing.
and an ideal selective glass is characterized by: This static aspect of the system is therefore
 opposed to precise calculations of the optical
= 0 et 0 soit = 1 for 0 < < c ,
properties of the rough surface and the many
= 0 et = 0 soit = 1 for > c .
models developed to determine them can only
(12-8)
do so based on drastic simplifying assump-
tions. We wont mention the formalism of such
An Example: Archaeoprepona
models since they are quite close to scatter-
These two cases are extreme, since the condi- ing models as far as their principles are con-
tion that is necessary and sufcient to decrease cerned. We will only focus on their qualitative
radiative loss, as dened by Kirchhoffs sec- predictions, which for lack of rigorous calcu-
ond law, doesnt imply any of these solutions. lations on butteries will enable us to evalu-
Butteries favoring direct absorption of solar ate the role of structures in buttery thermal
energy through their wings have generally de- balance. The most basic models allow to com-
veloped a third alternative. It is the absorption pare the reection R of a rough surface as char-
by the black areas situated at the wing base that acterized by the average height of rough-
provides the body with the most part of the in- ness and T the average period, to the reection
cident energy. The absorption is very high over R0 of a smooth surface of the same material,
the whole solar spectrumabsorption approx- for a given wavelength . The ratio R/R0 is
imates 98%but becomes insignicant in the a function of the only three parameters: , ,
Energetic Balance 139

will neglect structural scaleswhich, as we


have seen, sometimes contain pigments and
therefore contribute to the overall absorption
to focus on scales that are strictly pigmentary.
According to species, they can form two quasi-
continuous layers, one on the ventral side and
the other on the dorsal. Their structures as well
as their shape are characteristic. Contrary to
structural scales, pigmentary ones tend to be
lanceolate or multi-lobed, rather long and sub-
stantially overlapping.
Among Archaeoprepona menander, which has
been studied more fully from this point of
view, the pigmentary scales striae present a
very regular spacing of 1 m, while that of
counter-striae averages 300 nm with a higher
dispersion. They form rows of alveoles, ap-
proximately elliptic, over 1 m-high and open-
ing on the bottom on the scale alary membrane.
If one considers an average period T = 1, 3 m
and an average height = 1 m, the structure
leads to a ratio Rc /R0 , thus to a specular com-
ponent that is high beyond 7 m in the infrared,
and to a total reection, mainly diffuse, be-
low 30% in the visible. A signicant amount
of the incident energy is therefore trapped into
an alveolar structure which, as has been men-
tioned, consists of pigments that are efcient in
Figure 12.6. The denditric selective surface of a so-
this part of the spectrum although it does not
lar absorber. (a) Radiations of the short wavelength
radiations penetrate into the roughness where they take part in the infrared emission because of
are absorbed, while long wavelengths are reected. its weak absorption in this spectral range. Both
(b) rhenium dentrites. (after B.O. Seraphin, 1974) structures and pigments combine their efforts
in order to actually absorb over 95% of the in-
and T. As a rst approximation, one can dis- cident light over the whole visible spectrum,
tinguish a specular, or Rc coherent component even reaching 98% at 550 nm, the maximum of
and an incoherent Rd diffuse component in the solar spectrum.
the reection. As expected, the various mod- Infrared properties are even more remark-
els demonstrate that the coherent part has a able. Absorptivity reveals two strong peaks of
signicant inuence only in the long wave- absorption at 3 m and 6 m and remains
length range larger than the height of rough- more or less constant at a rather high value
ness ( > 5), which is not seen by the ra- approximating 0.4, up to = 50 m, which
diation anymore and thus hardly affect the marks the limit of our measurements. From
reection. On the contrary, the diffuse part the strict point of view of efciency, this sit-
prevails at small and medium wavelengths uation is not optimal. A rather high absorp-
( ), its intensity greatly depending on the tion hence, emissivitycovering the whole
ratio T/. The smaller the ratio, i.e., the deeper spectrum of the thermal emission is imply-
and the denser the roughness, the higher the ing substantial radiative loss. This is ignor-
absorption. Finally, if the average height is ing the double constraint subjecting the insect,
increased, the cut-off wavelength c shifts to- to raise its body temperature, yet not by too
wards infrared. What about the buttery? We much. As concerns this matter, this situation
140 Chapter 12 Thermoregulation and Spectral Selectivity

50C

0.5
40C

0.2 0.5 1 2 5 10 20 50

Figure 12.9. Absorptivity of the base of Archaeopre-


Figure 12.7. SEM image of Archaeoprepona pigmen- pona wings between 0.3 and 50 m, spectrum of solar
tary scales. irradiance and of the emission of a black body at 50
and 40 . The second absorption peak overlaps with
the emission spectrum at 50 , but it does not at 40 ,
leading to a quick change in the emission regime
approaches perfection and contains the outline according to temperature.
for a remarkable self-regulation phenomenon
of the radiative balance, allowing the stabiliza- emission spectrum, modifying the collector ef-
tion of temperature in the survival area of the ciency, and thus its temperature.
buttery. When after exposure to the sun, the tem-
The rst absorption peak, centered on 3 m, perature raises, the emission spectrum shifts
i.e., between the solar radiance and the thermal towards short wavelengths and overlaps the
emission spectra, is not involved in the radia- absorption peak. According to Kirchhoffs sec-
tive balance at these temperatures. However, ond law, this leads to a higher emissivity, hence
the second one, on the edge of the emission lowering the radiative efciency and decreas-
spectrum, plays a key role as a thermal regula- ing the temperature. On the contrary, when
tor. Its marginal situation is such that according there is no sun, the temperature is lower, the
to the wing temperature, it overlaps or not the emission spectrum shifts towards infrared and
T decreases

Efficiency
e increases

decreases
e increases
T increases

Efficiency
Figure 12.8. Reection, transmission, and absorp- increases
tion factors of Archaeoprepona black areasCu1 and
Cu2 baseshowing the strong spectral selectivity
of wings near the thorax and the combined action Figure 12.10. Temperature stabilization cycle of a
of reection and transmission that weakens the ab- buttery black wing. When temperature increases,
sorption in the infrared. the captor efciency decreases and the way round.
Energetic Balance 141

thus long wavelengths and the absorption peak the second case. The radiated energy is here
lies outside of the emission spectrum. In this proportional to the surface common to the ab-
case, the emissivity is lower, the radiative ef- sorption peak and the emission spectrum. At
ciency increases, and the temperature raises. 30 C, when the peak does not take part in the
The real efciency of this stabilization cycle is thermal emissivity at all, radiative losses reach
not easy to precisely determine when one does 275 W.m2 , but go over 500 W.m2 at 50 tem-
not know the temperatures that wings reach perature. The temperature of a buttery warm-
when exposed to the sun. One can still evalu- ing itself in the sun remains quite inferior to
ate the emissivity for the extreme cases of wing these extremes, yet the previous rough calcu-
temperature; for instance, 50 and 30 C. In the lations demonstrate that this auto-stabilizing
rst case, the peak totally overlaps the emission system can play a substantial part in the insect
spectrum, while it is outside of the spectrum in thermodynamic.
13
Vision and Colorimetry

Color Perception fundamental components, among which the


eye and the interpretation of the nervous mes-
The colored perception of an object is a com- sage by the brain is physiological and psycho-
plex and totally subjective phenomenon in- logical, and thus varies with the observer.
volving numerous parameters, physicochem- The light signal P() that reaches the ob-
ical as well as psychological. As has been servers brain can be described as a spectral
seen, butteries prove quite talented at draw- convolution of the light source irradiance E(),
ing attention or hiding themselves. We will the object reectancebutteries are usually
now consider the fellow creature, the preda- observed in reection, but the approach would
tor, or more generally, the observer viewpoint. be similar for transparent objects observed in
We will briey introduce the physiology of the transmissionand the spectrum of the sensi-
human eye and focus on the techniques to char- tivity of the eye to the different colors compos-
acterize colors, the stimulus, and the nervous ing the signal V():
message and its corresponding image in the
P() = E() R() V(). (13-1)
brain.
As has been mentioned, physiological and It is easy to quantify these various factors,
psychological phenomena involved in the per- especially the rst two. We will briey recall
ception of colors are complex and interwoven. the spectral characteristics of the sun, which
They are also specic: Insects, particularly Lep- is the source that has traditionally been used
idoptera, perceive colors in a slightly different to observe in vivo diurnal butteries. We
way from us, as is also the case of many of will here only overview the basic principles of
predators. Thus, the question why Butteries colorimetry.
are so vividly colorful, entails two answers de- As colorimetry allows us to dene a color
pending on the point of view. First, the point of in a nonambiguous way, as well as its pos-
view of physics: What are the physicochem- sible faithful reproduction, we will consider
ical processes, elements, or structures produc- the standardization of all the physiological
ing such a color? Secondly, the point of view and psychological parameters involved in vi-
of biology: What is the point of color as con- sual perception; in other words, the denition
cerns evolution? Indeed, if butteries survived of a standard source and average observer
with such colors, it implies that they are not the reference observer CIE 1931 (International
harmful to their evolution. Apart from these Commission for Lighting: 1931)to be a pre-
two points of view, it will only remains for us to requisite. This denition is based on the sta-
go into raptures on how ingeniously and del- tistical study of normal human vision, which
icately butteries use physical phenomena to is very different from that of a lizard or a
meet their goal. buttery.
It is difcult to precisely determine the per- The color perceived by an observer can be
ceived color of an object, as it involves three dened as the result of a measurement per-

142
Sources and Illuminants 143

pared to the whole energy emitted by the


source. An atomic emission ray that is quasi-
monochromatic has a purity of 1, while that
of white is equal to zero. All of this can actu-
ally be more or less independent of the spec-
tral composition of the wave coming in the eye,
since it does not perform a spectroscopy of the
wave but rather a non-univocal translation. We
will now describe more or less in detail each of
Figure 13.1. The color perceived by an insectivore the stages involved in the process of color per-
depends on the spectrum of the light coming from ception as illustrated in Figure 13.1. A distinct
the insect and reaching the retina photoreceptors,
i.e., both from the sourceusually solar, specular, chapter has indeed been dedicated to the study
or diffuse lightand from the wing reection co- of the optical properties of solids.
efcient. It can sometimes depend on the close or
distant environment of the insect, on the adaptation
state, and even on the predator experience. Sources and Illuminants
In nature, the universal source of light is the
formed by the eye, i.e., its various collectors, sun, the radiance of which is perceived directly
and interpreted by the brain. The result of or diffusely during the day and sometimes by
this process is the perception, by order of the reection on the moon at night. The monochro-
more obvious to the less, of three elements. matic distribution of solar radiance is complex,
First, a predominant color or shade, we tend to combining emission spectrum and ray spec-
say a dominant wavelength in physics (blue, trum. Beyond atmosphere, the spectrum some-
green, yellow . . . ); secondly, an intensity how looks like a black body at 5 800 K, yet it is
strong, intense, or weak lightdetermined by very altered when it reaches the earth ground
the amount of energy in the color spectrum. The after a more or less long tripdepending on
larger the intensity, the lighter the color. One the time and latitudeacross terrestrial atmo-
can thus establish a scale of intensities rang- sphere, one says traveled air mass, where a
ing from 0 (black) to 1 (white). And nally, a signicant part of energy is absorbed by ozone,
purity that determines the amount of energy mostly in the blue, and by water vapor in the in-
emitted to the predominant wavelength com- frared, before being scattered in the blue again.

Figure 13.2. The physical components of a visual stimulusarbitrary units. Light perception depends on
the spectral composition of the source, (a) on the reectanceor transmittance for transparent objectsof
the observed object, (b) and on the sensitivity of the various eye receptors to different wavelengths (c). The
corresponding brain representation is more complex to quantify, for instance, as coordinates within a color
spaceand is the goal of colorimetry.
144 Chapter 13 Vision and Colorimetry

Whatever the atmospheric conditions, the solar


spectrum always presents a lack in blue when
it reaches the ground.
Illuminants are theoretical radiations deter-
mined by their energy spectral distribution and
used as references for real sources. In our col-
orimetry calculations, we use a D65 illuminant,
the spectral distribution of which in the visible
and in the ultraviolet corresponds to a daylight
time with a color temperature of 6 504 K.

Colorimetry Notions
The Perception of Colors by Humans
Let us now study the aspect of the color percep-
tion process that is certainly the most complex;
the eye response V() and its interpretation by
the brain. Figure 13.4. Monochromatic solar illumination in
relation to the wavelength. Outside of the atmo-
The human eye contains two kinds of re-
sphere, the solar spectrum can be modeled as the
ceptors with very different properties. Cones, emission spectrum of a black body at 5800 K. At
which possess a high resolution yet weak sen- ground level, and according to the crossed air mass,
sitivity resolution power/capacity, are mainly it is slightly shifted towards red and presents nu-
concentrated in the retina central area: the merous absorption bands by the atmosphere com-
ponents. For an air mass of 4, spectrum intensity
fovea centralis. They are both photometric and
is substantially weakened and it is shifted towards
chromatic receptors allowing photopic vision red.
and colored diurnal vision. Mowing away from
the retina center, one then encounters rods
that are very dense at the periphery. Rods al- grey levels. The sensitivities of the two col-
low nocturnal scotopic vision. Those photo- lectors differ by more than three orders of
metric receptors are highly sensitive, but do magnitude (and their spectra are shifted); the
not perceive colors and only give an image in maximal sensitivity of cones peaks around
550 nm approximately(yellow-green), while
that of rods peaks around 500 nm (at the limit
between green and blue). Since the latter dont
take part in color perception, the denition of
the reference observer involves the response
curves of the various cones only. There are
three types of cones, each containing a pigment
presenting a maximal sensitivity in a specic
part of the spectrum: long (L), medium (M),
and short wavelengths (S). Sensitivity maxima
Figure 13.3. A color corresponds to a dominant are situated respectively at 560 nm in the red,
wavelength: 1 or 2 . This does not necessarily en- 540 nm at the yellow-green limit, and at 420 nm
tail that the wavelength is really present within the in the blue. These cones are sometimes des-
received spectrum, since two wavelengths can com- ignated by their predominant color: red (R),
bine and thus produce the effect of a third one (a). green (G), blue (B). These different types of
For a given dominant 0 wavelength, light intensity
can be more or less strong (A and B) and color more cones are not distributed evenly on the retina,
or less pure (C) according to the spectrum height especially the blue ones that do not represent
and width respectively (b). more than 2% of the fovea cones.
Colorimetry Notions 145

Figure 13.6. Cones relative absorption among hu-


mans and insectsnormalized responses. Among
humans, the green and red cones sensitivity curves
overlap to a large extent, which suggests they are
not mere chromatic captors. A red cone, for instance,
cannot differentiate between a weak stimulation in
the red and a stronger one in the green, because the
signals transmitted to the brain can be of equal in-
tensity depending on its sensitivity to the various
wavelengths. It is by comparing the signals coming
Figure 13.5. Inuence of the source (spectral) distri- from red and green cones that the brain can make
bution on colors; Morpho helenor 10 cm under day- a distinction. The same phenomenon occurs among
light air mass 1 - and illuminated by a tungsten insects with blue and purple cones. They also pos-
lamp color temperature (T 2580 K). The source sess a fourth type of cone that is sensitive to ultravi-
irradiances have been normalized. The blue radia- olet rays.
tion decreases to a large extent and the big spots
that slightly reect other colors are gray and weakly
bright. cerning its resulting color (Q), dened by the
following relation using only three variables
(R), (G), and (B):
Colorimetry Bases, Trichromy
(Q) = R(R) + V(V) + B(B) (13-2)
It is almost possible to reproduce all spectral
colors by superposing with adequate propor- where R, G and B are the amounts of primary
tions three monochromatic radiations; it is the colors contained in the mix, which also deter-
principle of additive trichromy. The colors es- mine the stimulus chromacity and its intensity.
tablished by the CIE are the red at 700 nm, the These two last quantities are usually separated
green at 546,1 nm, and the blue at 455,8 nm. by normalizing the trichromatic components,
The RGB system is based on them, but it is i.e., by dening the proportion of each primary
not the only possible combination. color in the mix.
Once the base has been determined, an ap- 

r = R (R + V + B)
propriate unit system has been adopted, a visi- 
ble electromagnetic radiationdetermined by v = V (R + V + B) (13-3)


its whole spectrum in physicscan be, con- b = B (R + V + B)
146 Chapter 13 Vision and Colorimetry

with:

r+v+b=1 (13-4)

The ratios r, v, b, are the trichromatic coordi-


nates of color. As they depend on each other, a
color can very well be dened by two coordi-
nates and its luminosity or intensity.
The same approach applied to the energy of
the monochromatic radiation E():

E() = r()(R) + v()(V) + b()(B) (13-5)

where r, g and b are the colorimetric coef-


cients, all the values of which for the various
visible wavelengths are the base of the colori-
metric functions of the RGB system. However,
certain very pure colors sometimes prove im- Figure 13.7. The 1931 CIE system X, Y, and Z compo-
nents. The Y curve is equal to the average sensitivity
possible to reproduce through additive combi- curve of human eye.
nations of the three key colors. That is why the
CIE suggested a new base in 1931: the (XYZ)
r The color depends on the dominant wave-
system CIE 1931. Thus, the colorimetric func-
tions x(), y(), and z() in the new system are length e , the value of which can be found
always positive and the y() curve precisely ts by prolonging the straight line starting in the
the human vision curve. As in the RGB system, white and intersecting the (x,y) color coordi-
the x = X/(X + Y + Z) and y = Y/(X + Y + Z) nates up to the spectrum locus. Prolonging it
reduced coordinates dene the chromaticity of in the opposite direction indicates the com-
the color in a (x,y) plane, known as chromatic- plementary color, if the intersection is on the
ity plane, whereas the Y value, which is propor- purple colors line, the color determined by the
tional to the visual intensity perceived, denes predominant wavelength e preceded by a
its intensity. For a given source dened by its sign.
r The purity is dened by the ratio OC/Oe .
spectrum E0 (), the chromatic coordinates of an
object with a reectance R() are as follows:

X = E0 ()R()x()d

Buttery Vision

Y = E0 ()R()y()d (13-6)

The anatomy of the extraordinary diverse pho-

Z = E0 ()R()z()d

tosensitive organs in animals is rather well
known today. Yet, it proves more tedious to
In a plane of equal intensity, the trajectory determine precisely the sensitivity of the cells
of monochromatic colors, purity equal to 1 (in- composing these organs. One can still mention
cluded in 380 to 780 nm, the theoretical limits several established general ideas on the vision
of human vision) describes a horseshoe curve, of butteries and of reptiles and birds, their
the spectrum locus closed by a straight line, the main predators.
purple colors line, characterizing the mixing be-
tween the two extreme colors of the visible
Anatomy of Compound Eyes
spectrum. The coordinates of the perfect white
(point O) are x = y = 1/3. Buttery compound eyes consist of a great
Any nonmonochromatic color is thus char- numberseveral thousandsof functional
acterized by its position C in the chromaticity units: the ommatidia, which are hexagonal and
diagram. arranged on hemisphere.
Buttery Vision 147

Figure 13.8. Chromaticity diagram. The color cor-


responding to a C point is determined by a domi-
nant wavelength e = 563 nm, a purity of 75%, and
a complementary color, that is here situated in pur-
ples (c = 563 nm). The Y tristimulus curve strictly
corresponds to the eye sensitivity curve, and the Y
value is directly proportional to intensity.

Ommatidia include three elements, listed


here from the exterior to the interior. First, the
corneal lens, a supercial dioptric device that
is a small plane lens either convex or bicon- Figure 13.9. Schematized representation of an eye
ommatidie by apposition.
vex. Second, a conic crystalline lens fringed by
pigmentary cells, which is extended by a third
element, the rhabdom, a deep nervous device
ter around 1, are little scattering in this spec-
fringed by 6 or 8 visual cells, the retinula. Each
trum. We here enter the typical eld of applica-
cell presents a nervous ber that transmits the
tion of effective medium theories. This rough
nervous message of the excited pigment to the
layer behaves, as concerns visible light, like an
brain.
homogeneous layer with an index ne that is
In apposition eyes of most of diurnal insects,
each ommatidium receives the incident light
under a very little solid angle only, oblique rays
being absorbed by pigmentary cells, which
also prevent light from entering next omma-
tidia. Resolution is high, but sensitivity is quite
low. The compound eyes of many Lepidoptera,
particularly Papilionidae, show on the corneal
lens surface a quasi-periodic structure consist-
ing of rather conic micro roughness, a hundred
of nanometers in diameter and 200-nm high. Figure 13.10. The corneule surface roughness create
The size of the latter being somehow inferior an index gradient serving as an anti-reection coat-
to visible wavelengths, with a size parame- ing at the air/corneule crossing.
148 Chapter 13 Vision and Colorimetry

Figure 13.11. SEM general image and details of Parnassius apollo compound eye.

intermediary between that of air (n0 = 1) and that could reveal the presence of the insect to
of the corneal lens (n 1,5). This index, ap- predators. These devices, known as moth eyes
proximating 1,2, can be evaluated thanks to structures, are also used in the space industry.
Maxwell Garnetts theory or through its ap- How images are formed through such an op-
proximation to neighboring indices. This effec- tical device is complex and varies according
tive layer behaves like an anti-reecting coat- to species. The apposition eyes, in each com-
ing, both allowing to reduce optical loss at the pound of which rays enter and give an inde-
ommatidium entrance and to avoid reections pendent signal, are distinguished from the su-
perposition eyes of generally nocturnal species,
in which rays can be combined on the same re-
ceptor. These eyes do not allow any focalization
but are highly sensitive to motion. A fact that
concerns us more is that butteries can perceive
colors well and that their sensitivity spectrum
is shifted towards ultraviolet as compared to
Figure 13.12. Comparison between the photorecep-
tors sensitivity spectra of humans, Lepidopterons, ours. They can therefore perceive on their fel-
and Birds. The buttery spectrum goes far beyond low creatures motifs that are invisible to preda-
that of its main predators in ultraviolet. tors.
Buttery Vision 149

If the range of their spectral sensitivity is


rather well known, the capacity of distinguish-
ing colors among insects has only recently been
established and precisely only on a small num-
ber, mostly bees. Studies applied to other cate-
gories are incomplete, not introducing any dra-
matic change. It is accepted that these species
can perceive colors, yet the perceived color re-
mains to be determined.
The majority of the observed Lepidoptera
show four types of receptors (Figure 13.6) with
rather wide sensitivity band (around 100 nm)
respectively centered in ultraviolet (350 nm),
blue (450 nm), blue-green (500 nm), and green- Figure 13.13. Axial or longitudinal pigmentary mi-
yellow (540 nm). At 350 nm, the UV peak is al- grations along the rhabdome enable to adapt to var-
ways narrower than the peaks observed in the ious intensities. In the case of axial migrations, pig-
visible. The visibility spectrum is thus on aver- ment granules are lining the rhabdome inner sur-
age shifted towards ultraviolet by around 100 face and absorb light during the different reections.
They get away from the surface when the intensity
nm in comparison to the human spectrum, a lit- decreases.
tle less in comparison to that of birds, reptiles
or arboreal amphibians, their main predators.
This gives them many advantages, among transmitted and retro-scattered lightand the
which for some of them is the ability to polarization direction is always perpendicular
displayin spite of a similar cryptic or mimetic to the scattering plane. In this way, light coming
aspect in males and females in the visiblea from a clear blue sky, is always partially polar-
strong sexual dimorphism in ultraviolet, thus ized, excluding certain neutral pointsBabinet
allowing them to better recognize each other
without any risk.
Some owers, which are pollinated by in-
sects, show also in the ultraviolet very bright
distinctive marks due to the presence of
avones, which attract insects searching for
nectar.

Perception of Polarized Light


As has been seen in the previous chapters,
whatever the physical phenomenon involved,
colors produced by structures are partially po-
larized. They are even more polarized when
incident light is polarized. There are two im-
portant aspects involved in this problem: (1) Is
natural light polarized? If it is, the polarization
effects generated by the insect are amplied.
(2) Are insects sensitive to this effect; in other
words, can they distinguish between different
polarization states? Figure 13.14. Isopolarization map of the sky at sun-
rise (S). The two nonpolarized areas above the sun
(1) As has been seen in Chapter 10, the scat- and its image (G), are respectively known as Babinet
tering of a wave leads to a partial polarization point (B) and Aragon point (A). Arrows show the
of light, except for the direction of incidence polarization direction of light.
150 Chapter 13 Vision and Colorimetry

point and Arago pointthat are not always vis- prehensive, it is accepted that a great number
ible. of them can perceive polarization states. This
This has two important consequences. First, perception is based on the variations in absorp-
when the sky is clear, diffuse light falling on tion of the electric eld with its orientation. In-
an object or an insect is polarized mostly per- deed, it has been discovered that the absorption
pendicularly to the scattering plane. The po- of some materials depends on the polarization
larization rate can reach about 70% at 90 from states of the incident light, either rectilinear or
the sun direction. According to the orientation circular. Known as dichroism, the phenomenon
of the insect or of the structure in relation to is caused by the shape of absorbing molecules,
this plane, polarization effects by reection on which are either elongated in a single direction
wings can be amplied, thus increasing the or in spiral. A linear dichroism has indeed been
contrasts. discovered in the visual cells of rhabdomes in
This also entails that a measurement of sky some insects, and orientation maps have even
polarization can reveal the direction of the sun, been established. It has thus been proved
even if is hidden. It has been established that even if their meaning remains obscure
bees actually use this property to nd ones that all polarization effects generated by the
bearings, and perhaps also by some big mi- complex structures previously mentioned
grants like Danaus plexippus. can be perceived by most insects and there-
(2) Even if the study of insect sensitivity to fore represent one more vehicle for visual
light polarization is still far from being com- communication.
Conclusion

Since its creation, the laboratoire doptique des Let us rst describe the traditional struc-
Solides, a CNRS-Pierre et Marie Curie Univer- ture of a fabric. Its basic component is ber
sity unnite mixte, has focused on the study synthetic in the present case. It is a long ber,
of optical properties of inorganic solids, rst usually about 10 nanometers in diameter, com-
on metals, and later on semiconductors, di- posed of various more or less long polymer
electrics, and inhomogeneous materials. Noth- chains, the nature and arrangement of which
ing to do with butteries indeed. And yet, determine the mechanical properties of the
the further we study wings, the more simi- ber. It is on this ber that the optical struc-
larities we nd. Or rather, we realized how ture generating iridescent colors are produced.
much butteries were ahead of us concern- These bers are then gathered in a varying
ing optical phenomena. They in fact proved to numberusually 60and form the thread that
conceal all that concerned us: dielectric multi- is woven or knitted to create fabric. During
layer, scatterers, selective materials, disordered the various stages, one can notice that the ba-
mediaand that only concerns wings: all sub- sic element supporting the very irregular op-
jects that we were studying. In this way, but- tical structure is subjected to numerous dis-
tery wings became quite naturally part of our tortions at extremely different scales. The ber
research. Materials are new, yet structures are curving rays, for instance, can be measured in
traditional. Our research resulted from a pas- tens of micrometers when twisted into a thread,
sion dating back to my childhood and a denite and some millimeters once woven. Finally, the
industrial request. creases in the nished garment present distor-
Within the framework of a prospective re- tions measured in centimeters.
search on the fabric of the future, we were The main difculty is then to model, or at
asked about the origin of iridescence among least evaluate, the evolution of the optical prop-
butteries and its possible application to tex- erties at these various scales. Physical phe-
tiles. The answer to the rst question is the topic nomenon involving elements occurring at dif-
of this book. It didnt reveal more complex, yet ferent scales are among the most complex.
it proved to be more diverse than expected. In fact, it is to them that one is confronted
As concerns the application to bers, and even when one wants to evaluate the visual effects
garments, represents a real challenge, even in- of a garment, the iridescent colors of which
dependently from any technical creation. It is generated at the smallest scale and imposing a
worth mentioning the method. As the intro- strict orderare modulated by local entropies
duction was dedicated to the buttery, so the of increasing scales. The solution we chose is
conclusion has to be. One thus starts by den- directly inspired by Urania scales. It consists in
ing precisely the optical properties of a plane distributing over the ber as it is being woven,
multilayer. It is then rather easy to optimize the a certain number of thickness and index lay-
system, based on the materials available for tex- ers corresponding to a given color. Calculations
tiles, in order to obtain a greater iridescence. show that this number actually does not need

151
152 Conclusion

Figure C.1. Transition from a plane multilayer to


a cylindrical multilayer. One actually models a
polygonal-section ber. If the number of facets is big
enough, their width is inferior to the eye resolution
limit.

to be excessive. Manufacturing process and


other various constraints impose a minimal
thicknessaround 100 nmand specic ma-
terials. The optical index that are available are
included between 1.4 and 1.7. One can there-
fore determine the multilayer parameters
thickness and indexwhich generate an op-
timal iridescence centered on a given color.
By modifying these parameters as well as the
number of layers, one can increase or decrease
as one wishes the luminosity, color purity, and
iridescence amplitude. It proved that a mere
piling of two or three layers can produce the Figure C.2. Modeling of a thread. Fibers that are ran-
effect that is wanted. domly disposed in the thread are arranged as planes
in the model. These inhomogeneous planes are
Difculties start at this point. The rst is then assimilated to continuous planes.
geometric and consists in passing from the
plane multilayer on which calculations were
based, to the same multilayer arranged on a One then has to determine and character-
cylinder, i.e., the ber. The simplest way is ize the effects that can occur when bers are
to assimilate/compare the ber circular sec- twisted to form a thread. The resulting disorder
tion to a polygonal section. The ber thus and the system geometry require drastic sim-
doesnt consist of a continuous cylindrical sur- plications, the consequences of which are dif-
face anymore, but of facets. For a 11,5-m total cult to evaluate. With an approach quite sim-
diameterber + multilayerwe divided the ilar to that adopted to model Morphos optical
section into 5 angular sectors; that is 72, 0,5-m properties, one can model the thread as a regu-
wide facets. If one considers an isotropic dif- lar piling of bers lining on one plane and sepa-
fuse lighting, one can determine the light spec- rated from the other pilings by a very thin layer
tra reected into a given direction by each facet of air. Calculations thus predict a dramatic in-
and also the characteristics of the observed crease in intensitywhich has yet to be some-
color, by taking into account shadowing ef- how relativizeddue to the perfect piling of
fects and possible multiple reections proceed- bers, which is produced only locally in a real
ing from neighboring facets. Simulations reveal thread. If the amplitude of iridescence is little
that passing from plane to cylinder doesnt af- affected from one scale to the other, the second
fect much the iridescence amplitude, nor the predicted phenomenona signicant decrease
purity of the resulting colors. in color purityis less encouraging. The effect
Conclusion 153

Figure C.3. Iridescence variations during the rst


stages of modeling of cloth. Iridescence amplitude
hardly varies, but the purity systematically de-
creases, especially as concerns central colors. Figure C.4. From the structured ber to cloth, pass-
ing through meshes . . .

of the weaving and knitting of threads has not


been modeled, yet one can also expect a loss in
purity as far as it is concerned.
Are these approximations relevant, and is it
worth it in the end? In order to answer such
questions, one needs a reference, which nut-
teries can provide. Indeed, the comparison
between wing and garment is more relevant
than it seems. One thus nds quite precisely
the same size scales and the same structures in
both of them. As has been seen among Morphos,
for instance, the optical structure is present on
striae, covering the scale surface in hundreds,
the typical size of which approximates some
micrometerslike ber. Among other butter-
ies, the optical structure is situated on the very
scales surface. Like in a thread, scales average
100 m and are regularly distributed on the
alary membrane just like in weaving. More-
over, the membrane can be measured in cm2
and includes various creases, undulations, and
irregularities typically measured in millime-
ters. Since the structures and size hierarchies Figure C.5. Steps and scales are similar among but-
are similar in both cases, one can expect that teries, from their structured scales to their wings.
colored effects will also be similar.
154 Conclusion

When one compares the experimental iri-


descence curves of a Morpho and the calcu-
lated ones of an iridescent fabric modeled as
above, they appear quite similar. Buttery col-
ors which seem so bright to us, are in fact not so
pure. Therefore, a purity equal or little inferior
to that of butteries would be quite acceptable
for fabric.
The technical difculties have been partially
overcome in Japan, for instance, where such Figure C.6. Evolution of the polarization of a wave
iridescent fabrics are starting to be produced. subject to a double reection in a groove under an
incidence close to the Brewster incidence. The emer-
Another question and another debate would gent wave can be Transverse Magnetic polarized
be to what extent one would be ready to wear over more than 80%.
these changing colors.
Another example of biomimetic transfer is
the production of colored effects due to polar- colors, papers, watermarks, and holograms.
ization. One of the main applications would The second level of protection requires quite
be the marking of bank notes, cards, and so simple equipment and concerns shopkeepers
on in order to identify them and guarantee and cashiers. It implies concealed properties,
their authenticity. The inspiration now comes such as ultraviolet ones or even polarization
from papilio and Cincideles, which is to say in- ones like in the case concerning us. Level 3 pro-
sects presenting concave interferential struc- tection involves all the physicochemical prop-
tures on their wings. As has been seen, in such erties of the bank note, which can only be
structures, the interferential color observed is demonstrated in a laboratory by using cumber-
a compound color resulting from the mix of some equipment and therefore only concerns
the little polarized one produced by the basin central bodies.
bottoms and that generated by the inclined Once their symmetry broken, the effects pro-
edges, which reveals highly polarized after be- duced by Papilio and Cincidele structures can
ing polarized twice. In this way, there should bring an original answer to the problems that
be a way of modifying color by suppressing the rst two levels entail. One can on the one
the second component thanks to a mere polar- hand generate bright iridescent colors fullling
izer. This is almost impossible to perceive on the level 1 protection requirements and on the
insects because structures are generally sym- other hand, create chromatic effects depending
metric. There are as many horizontal as ver- on polarization that correspond to the level 2
tical edges and the colors reected by these protection.
edges are polarized perpendicularly to one an- One can roughly imagine the process as fol-
other, which implies that there is no global po- lows. Colored motifs are composed of grooved
larization of the component. Yet, one just has and at areas that are juxtaposed and form
to break the symmetry by producing grooves a drawing or a sign. All the areas contain in
rather than basins in order to maximize the ef- their thickness a multilayer structure produc-
fect. ing a high iridescence. As they are not illumi-
One must have a few notions on bank note nated under the same incidence, the grooved
protection in order to fully understand this and at areas generate different interferen-
approach. Each bank note consists of a cer- tial colors. Under near-normal incidence, the
tain number of elements ensuring protection light reected by at areas is hardly polarized,
at three different levels corresponding to three whereas that of grooved areas is highly polar-
types of users. Level 1 protections must make ized. The groove period must be inferior to the
anybody in the street or in a shop be able to rec- eye resolution in order that these areas appear
ognize without any equipment the authenticity perfectly homogeneous. For an observer stand-
of the bank note. This involves various motifs, ing some 20 cm away from the bank note, it
Conclusion 155

requires a width of 60 m, which one can eas-


ily create.
One can then calculate the colorimetric prop-
erties of these areas, as well as their irides-
cence and polarization states under various
incidences by considering the shadowing ef-
fects proceeding from grooves. The approach
is quite similar to that adopted for fabric; so
we wont describe it further. Let us only men- Figure C.7. A colored effect linked to the polariza-
tion that one should not expect to obtain a to- tion of one of the components. By suppressing a large
tally polarized light with such a process and part of the blue component with a polarizer, one can
make the general green color turn yellow.
that the contrasts between areas wont be max-
imal. The Brewster angle, which, as has been
seen, depends on the two dioptric materials
ries in this eld! And our brief remarks out-
index, is not determined by one multilayer
side the main track, especially compound eyes,
anymoreas indices and incidences vary from
showed that such elements have many sur-
one layer to the other. The p polarized light still
prises in store.
presents a minimum for a pseudo Brewster
Yet above all, butteries proved to be great
angle for which the reected wave is highly
a vehicle for teaching physics, and as concerns
polarizedsometimes up to 80%. Different vi-
us, optics. This new and quite unusual theme
sual effects can be imagined thanks to this
device. Here are a few examples. The one re-
sembling Papilios and Cincideles the most is the
succession of grooves and alternated planes of
the same width. This results in a compound
color, one of the components of which can be
suppressed to modify the color by using a po-
larizer. The motif color can thus vary from
green to yellow by either suppressing or letting
enter the blue component reected by multi-
layer grooves that reect yellow under normal
incidence. Another original method consists in
using only grooved areas that are perpendic-
ular to one another. Without a polarizer, they
show the exact same color, but when perpen-
dicularly polarized, only one or the other ap-
pears, thus making concealed motifs emerge by
observing them with a polarizer.
Let us now put aside the potential ap-
plications of the present research and intro-
duce a completely different matter. In the past
and more or less successfully, butteries in-
spired poets, and now engineers. In fact, after
striving to invent everything from scratch for
decades, we are now coming back to nature to
show us new paths. We were in the present
book only concerned with optical properties Figure C.8. When one uses grooved areas only, they
present the same color under natural light and it is
of wingsgratings, Bragg mirror, diffractive impossible to distinguish motifs. They can become
optics, spectral selectivity, disorderand but- apparent by using a polarizer. TE and TM are relative
teries proved to be real high-tech laborato- to the central motif structures.
156 Conclusion

was started prudently, yet pupils quickly took blue wing (it immediately turns green, then
an interest in itfrom middle school interns, blue again when dry). The same experiment
who thought it was impossible to work seri- performed on a blue Argus results in the disap-
ously with this material, to high school sci- pearance of blue without the green stage! The
ence juniors and seniors, who discovered (bet- three main sources of colors in the living world
ter late than never) a totally different aspect of have thus been experimented. Then, a group
physics. They would thus attend conferences observes and carries out the same experiments
on butteries, sometimes several hours long, by using an optical microscope, modifying in-
without showing impatience, and I would even cidence, while another group researches the
say with a certain enthusiasm . . . and at the various origins and properties of light. These
end, they would realize they had actually classes went out, to discover in museums how
studied physics for four hours. The interdis- it works among birds, shells, stones, and so
ciplinary study of butteries proved a success on. When they realized that iridescence could
in high school, and a great way of introduc- not be generated by the striae they had ob-
ing physics. We wished these bridges between served with the optical microscope and that
different subjects could be implemented in they had to change the scale once again, we or-
universitiesthis modest rst experiment in ganized several scanning electron microscopy
high school being quite encouraging. This is sessions at the university. The resulting pic-
in fact worth describing it. Launched between tures were then studied in high school: to have
biology and physics via butteries, the ex- a class count the diffusing particles of a Cab-
periment could very well be applied to any bage Pieride is worth any image-processing
other organism that is a little popular accord- software . . .
ing to meshells, beetles, and, during spring In the end, these pupils who tend to con-
months, grasshoppers, maybugs, or owers sider physics as one of the most boring subjects
and between any subjects: thermodynamic, discovered physics and basic optics notions
mechanics, chemistry, mathematics, and even via a more appealing subject like biology. No-
arts or music. I had the good fortune to meet tions that they will be less reluctant to study
a biology and a physics teacher who agreed to further in class. For instance, they saw a
adapt classes and dedicate some time to imple- wavelength and evaluated its length for the
ment the project together. It entailed changing rst time. Better still, the experiment showed
the traditional schedule of one-hour physics them the advantages of interdisciplinary stud-
class and a break followed by a one-hour bi- ies, how interfaces enrich one another, but also
ology class. So, we asked pupils to resolve a how jargons could represent a barrier between
complex interdisciplinary problem: How and two communities. May they persevere in this
why do butteries present such colors? This re- path, and always remember this new butter-
quires quite simple experiments that anybody y effect. And may this book guide them. I
can carry out: a piece of red wing is dipped into would be very proud. It would be my best
acetone (nothing happens), a piece of Morpho achievement.
Subject Index

Cetonia aurata, 44 Morpho menelaus, 91, 94, 95, 101, 102


Charaxes jasius, 19f Morpho sulkowskyi, 105
Chrysiridia madagascariensi, 72, 73 Morpho zephyritis, 101, 102, 105

Danaus plexippus, 150 Notasacanta dorsalis, 55


Dynastes hercules, 55
Ornitoptera priamus poseidon, 18f, 30f
Eupholus humeralis, 115f
Eupholus sulcicollis, 115f Papilio blumei, 54f
Papilio paris, 80f, 81f, 82
Hoplia argentea, 84 Papilio ulysse, 53f, 82
Parnassius apollo, 20f, 24f
Melitea phoeba, 15f Phaleanae Biston betularia, 09
Morpho cypris, 101, 102
Morpho menelaus, 89 Urania leilus, 72, 73

157
General Index

2-dimensional structures cathedrals, 64


type butteries with, 88 cephalopodes, 55
Cetonia aurata, 44, 48
adult buttery circular polarization, 82, 85
body of, 19 Charaxes jasius, 19f
alary cells chirality, 43
median axis of, 26 chitineou layers, 31
alary membrane, 31, 33, 92, 93, 95 cholesteric phase, 43, 44
amorphous phase, 67. See also development of butteries chrysalis development stage, 15, 30
androconia scale, 34 Chrysiridia madagascariensi, 72, 73
angle of incidence, 53, 71 Cicindelidae, 40, 44
aniline circular polarization, 82
thin lms of, 03 Cithaeria menander, 30f, 31f
anisotropic material, 59 Coleoptera, 40, 41
apetetic colors, 10 cuticle of, 40
archaeoprepona, 27, 138, 139 life cycle of, 41
archeoprepona, 36, 53, 131 scales of, 41
arthropods, 40, 42 study of, 56
articulated legs colors, insects, 26
pair of, 18 color perception, 142143
atomic force microscopes, 56 colored effects, 77
auxochrome, 128 colored motifs, 28
cryptic colors, 135
batesian mimicry. See camouage palette of, 07
biliary pigments, 130, 133 protection provided by, 08
biotypes, 07, 23 compound eyes
innite diversity of, 07 anatomy of, 146
Biston betularia, 09 crystalline material
blue pigments, 04, 53 intra-band transitions of, 04
brewster incidence, 58, 59, 77 Curculionidae, 48, 112, 115. See also
butteries Coleoptera
absorption of solar energy by, 06 cuticle
biomimetic applications of, 05 structure of, 41
development of, 15
diurnal and nocturnal, 14 Danaus plexippus, 150
high reproductive rate of, 04 de effective electric displacement, 63
interior temperature of thorax, 08 dephasor, 85
optical phenomena in, 42 depolarizing factor A, 65
repelling taste of, 10 development of butteries, 15
vision, 146 dielectric function, 63, 64
diffraction
camouage, 4, 8, 9, 10, 15, 55 hypotheses of, 03
caterpillars, 9, 15, 18, 40 diopter, 68
ingestion of soot by, 09 Dynastes hercules, 54, 55

158
General Index 159

ecologic context, coloration, 07, 08 leucopterines, 55


edible butteries, 08 linear polarization, 57, 59
effective index, 63 Lycaenidae, 03, 133
electromagnetic eld
vector nature of, 58 mandrills
electronic microscopes muzzle of, 03
scanning and transmission, 56 Maxwell equations, 57f
elliptic polarization. See circular polarization Maxwell, J. C., 56
elytron wings, 71 melanins, 129
embryonal stage, 15 melanism phenomena, 09
endocuticle, 44 Melitea phoeba, 15f
Eupholus humerialis, 115f metallo-protein pigments, 53
Eupholus sulcicollis, 115f metals
exocuticle, 83f colloidal suspensions of, 64
metamorphosis, 01. See also development of
uorescence phenomena, 49 butteries
frenulum, wings passive device 19 earth to ether, 01
frivolous symbolism, 01 passage, 01
Michelson, A. A., 2, 3, 56, 62, 86
goniochromic structures, 54 Mies treatment, 119
Grapium weiskei, 16f mimicry
basic principles of, 11
haemocyanin, 53 modalities of, 10
haemolymph messages, 22 mullerian mimicry, 110
hard sciences, 01 Morpho cypris, 101, 102
hatching, 26 Morpho diffraction maps, 2, 4, 36, 38, 39, 50, 51, 55, 65, 72,
Hebonia glocippe, 52f 88
helicoidal structure, 42 Morpho menelaus, 89
heterotherm insects, 08 Morpho menelaus, 91, 94, 95, 101, 102
Hoplia argentea, 84 Morpho sulkowskyi, 105
hydrochromic structures, 55 Morpho zephyritis, 101, 102, 105
Hymenoptera, 14 Morphos Adonis, 108
mullerian mimicry, 110
imago, 15, 18 multilayer, structural scales, 49
incidence various Types of, 71
angles of, 03
industrial melanism, 09 N-acetylglucosamine
insects monomers of, 42
classication of, 05 Notasacanta dorsalis, 55
electromagnetic properties of, 06 nymph, development stage
photonic crystals, 115 diapause, 18
standard bearers of, 04 shedding, 30
integument
surface of, 55 oligopod, 40
interstriae spacing, wings, 38 olympic gods, 01
intra-band transitions, 04 ommochromes, 131
ionic crystals, 60 one-dimensional crystals, 66
iridescence, 54 optics
isochromy, 10 optical index, 63
optical properties
Jones vector equation, 59 study of, 60
optical recognition, 07
Kosciuscola, grasshopper, 55 Ornitoptera priamus poseidon, 18f, 30f
Kramers-Kronig equations, 31 orthonormal reference, polarization, 59

Larval/nymphal stages, 15 Papilio blumei, 54f


Lepidoptera, 14, 16 Papilio paris, 80f, 81f, 82
pigments, 129 Papilio ulysse, 53f, 82
protective strategies of, 12 papiliochromes, 132
160 General Index

papiliochromis, yellow, 27 scattering structures


Papilionidae, 79, 81, 134, 147 butteries with, 123
parasematic colors, 10 selective reection hypothesis, 03
parasitism, 12 sematic colorations, 10
Parnassius apollo, 20f, 24f sexual recognition, 08
peduncle, 48 solar spectrum, 08
pheromones, 07 solid state optics
photonic crystal, 65 notions of, 06
physico-chemical origins, 10 specular reections, pigments, 49
Pieridae, 67, 100, 109 structural origin, colors
pigment, 4, 7, 16 defenders of, 03
scattering with, 04
pigmentary effects, 93 tetrapyrolique pigments, 53
pigments thermochromic structure, 55
interference/diffraction effects, 51 tiger beetles. See Cicindelidae
Plan Diopter, 68 tillards system, 24
plane grating, diffraction, 86 tissues
polarization, 83 differentiation of, 15
matrix approach to, 59 reconstruction of, 15
polarization effects, 114 toxic compounds, 07
polarization of light, 149 tropical butteries, 53
measurement of, 150
polaroid lms, 56 ultraviolet
polymorph mimicry, 13 iridescence in, 109
polynomial expression, 62 Urania leilus, 72, 73
polypod, insect larvae, 40 Uraniidae, 14, 72, 73
predators, 13
Prepona, 72, 78 vegetal cellulose, 42
procuticle, cuticle structure, 41 venation, 22
pseudo-sematic, colors, 10 visible spectrum
pterobiline pigment, 134 splitting of, 04
pterotecae wings, 30 visual camouage
pupal stage. See chrysalis development the art of, 08
stage
Walter, B., 03
Raleigh, Lord, 03 wings
relative immunity, 12 anterior wings, 10, 19, 31
rotator structure, 60 extensors and abductors of, 08
macroscopic level, 22
saturnides, 14 odoriferous roles, 29
scales posterior wings, 10, 13, 14
morphologies, 29 radiation qualities of, 06
nomenclature attempt, 29
structures, 29 X-chrome, 53
Scarabeidae, 48, 84 0 frequencies, 61

You might also like