Hybridization of the Habitat

Author(s): Edgar Anderson

Source: Evolution, Vol. 2, No. 1 (Mar., 1948), pp. 1-9
Published by: Society for the Study of Evolution
Stable URL: http://www.jstor.org/stable/2405610 .
Accessed: 10/03/2011 08:02

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=ssevol. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Society for the Study of Evolution is collaborating with JSTOR to digitize, preserve and extend access to
Evolution.

http://www.jstor.org
 EVOLUTION
INTERNATIONAL JOURNAL OF ORGANIC EVOLUTION
PUBLISHED BY

THE SOCIETY FOR THE STUDY OF EVOLUTION

Vol. II MARCH, 1948 No. 1

HYBRIDIZATION OF THE HABITAT

EDGAR ANDERSON

Missouri Botanical Garden, St. Louis 10, Missouri

Received November 7, 1947

It has been the experienceof mostbiol- ics (see for instanceEast, 1913, 1916)
ogists that hybridizationbetweenspecies and experimentaltaxonomy (Clausen,
is rare in nature. Many biologistsen- Keck, and Hiesey, 1946), and the experi-
counterinterspecific hybridsin the field ence of numerousplant breeders. The
so rarelyas to doubt if theyreallyoccur latter,by far the largest of these three
thereat all or else findthemunder spe- bodies of evidence,is not too accessibleto
cial circumstances(Epling, 1947) which most scientistssince it has to be dug out
raise serious doubtsas to the importance piecemealfromsuch compendiaas Reh-
of hybridization undernaturalconditions. der's Manual of Trees and Shrubs(1940).
Aftera series of investigations(Ander- A criticalsummaryof all this evidenceis
son, 1936b, 1936d; Anderson and Hub- badly needed. Lacking one, the evidence
richt,1938; Andersonand Turrill,1938) forthehigherplantsmaybe roughlysum-
it has been myown experiencethatclear- marized as follows: Well-differentiated
cut out-and-outhybridsare seldom met speciesof the same genusmayor may not
with, even when a deliberatesearch is be interfertilewhen tested experimen-
made for them,and that hybridswarms tally. On the whole it seems to vary
of bizarrerecombinations are found,if at with the genus. There are certaingen-
all, onlyunderpeculiarcircumstances. era in which interspecifichybrids are
It was once the commonopinion (Zir- difficult to make and are sterile. There
kle, 1935) that this lack of evidenthy- are others,equally exceptional,in which
bridizationwas caused by the sterilityof the widest possible crosses within the
interspecifichybrids. Experimentalevi- genus will yield fertile or semi-fertile
dence has not confirmedthis judgment hybrids (Anderson and Schafer, 1931).
and modern advocates of reproductive The yellowtrunmpet Narcissi are so com-
isolation (Mayr, 1940) as a species cri- pletelyfertilewiththe flat,white-flowered
terion have had to phrase their defini- Poets' Narcissi that the whole business
tionsto permitsemi-fertility betweendis- of supplyingnew gardenhybridshas been
tinctspecies. For the higherplantsthere founded on it (Calvert, 1929). The
is an impressiveamoun-tof experimental Poets' Narcissi (in themselvesa whole
evidenceon this pointthoughit is widely group of species and sub-species) are so
scattered. There are the papers of the interfertile withthe speciescomplexmak-
earlyhybridizers(Focke, 1881; Roberts, ing up the long-crownedyellowdaffodils
1.929;Zirkle,1935), muchworkin genet- that it has been possible to accomplish
EVOLUTION 2: 1-9. March, 1948.
2 EDGAR ANDERSON

such recombinationsas the transferof It is usually more vigorous than either

the deep red-orangepigmentfrom the and more robustin nature.
rim of the tiny central eye in a Poets' 2. The secondhybridgeneration,while
Narcissus to the brilliantflaringorange on the average intermediate, is extremely
trumpetof such modern"red trumpets" variable. Usually no two individualsare
as the variety"Fortune" (Anderson and alike. The variationis usuallybewilder-
Hornback,1946). The commonestcondi- ing,but it can be shownto have a general
tion amongthe higherplantsseems to be trendfroma fewindividualsmoreor less
that in which crosses within a species like one of theparentalspecies,to a great
group are easy to make while intergroup bulk moreor less like the firstgeneration
crosses are difficult or impossible. hybrid,to a very few more or less like
If, therefore, we base our explanation the otherparentalspecies.
of the rarityof hybridsunder natural 3. The facts with regard to the back-
conditionson the experimentalfacts we crosses are equally well establishedbut
can make the followingsummary:1. In unfortunately have been so littlestressed
a minority of the cases whererelatedspe- in modern times thattheyare not as gen-
cies occur near each other,theyare com- erally familiarto biologists. If the first
pletely intersterile. 2. In certain cases hybridgenerationis crossedback to either
hybridscan be formedbut are sterile. parent, the firstbackcross is made up
3. In a surprisingnumberof cases the ex- targelyof plantswhichresemblethe spe-
perimentalevidence shows that the spe- cies to whichtheyhave been backcrossed.
cies can be crossed readily,in the breed- If we take one of these and cross it back
ing plot, but no hybridsappear under a second time to the same parent, not
natural conditions. 4. In many cases onlywill the seedlingsresemblethis spe-
which have been carefullyinvestigated, cies verymuch indeed but some of them
hybrids,though usually rare, do occa- may be nearlyor quite indistinguishable
fromit. In all the cases with which I
sionallyoccur but leave no apparentde-
am personallyfamiliar,many,ifnot most,
scendants except under unusual condi-
of the firstand second back-crosses,if
tions.
found in nature, would by taxonomists
This experimentalevidence,such as it
be acceptedas varietiesor slightlyaber-
is, justifiesthe generalizationthatspecies
rant individualsof the species to which
maintainthemselvesas recognizableunits
theywere back-crossed(see fig.1). Few
even wheretheyare interfertile and even taxonomists,even those specializing in
when thereis a considerableopportunity that
group of plants,would even suspect
for themto hybridize. Why should this that
many of these back-crosseswere of
be so? Beforewe can answer the ques- partiallyhybridancestry. For the genus
tion we must firsthave clearly in mind Apocynum we have a detailed experi-
the known resultsof such hybridization mental record of back-crossmorphology
under experimentalconditions. If we and the taxonomicreactionto it (Ander-
ignore the complicationsdue to poly- son, 1936b).
ploidyand otherimportantbut more-or- In the genus Tradescantia,by rigidly
less specializedphenomena,the usual re- experimentalmethods(Anderson,1936a,
sults of hybridizationare readily sum- 1936d; Andersonand Sax, 1936; Ander-
marized. They were well establishedby son and Woodson, 1935), includingthe
theearlyhybridizers and have been abun- productionof experimentalbackcrosses,
dantly confirmed by modern genetic it has been possible to demonstratethat
research. the principal result of hybridization,in
1. The firsthybridgenerationis inter- thosecases whereit didoccur,was a series
mediate between the parents and is as of such backcrosses to one or to both
uniformas they are, or even more so. parents. To thisphenomenonI gave the
 HYBRIDIZATION OF THE HABITAT 3

N. LANGSDORFFII

X,z>
N. ALATA

FIRST BACK-CROSS

FIG. 1

nameintrogressive hybridization(Ander- crosses could not be recognized as of

son and Hubricht,1938), since it pro- mongrelorigin purely by their appear-
vided a mean by which elementsin the ance. It is thereforeclearly indicated
germplasmof one species might intro- that gene flow fromone species to an-
gress into the germplasmof the other. othermay go farbeyondany pointwhich
The chief result of hybridizationunder could be detectedby ordinarymorpholog-
such conditionsis the enrichment of vari- ical techniques. We shall not be able to
ation in the participatingspecies. Such assess the real importanceof introgres-
hybridization is crypticand only by very sion until we can study geneticallyana-
specializedtechniquescan we measureits lyzed species in the fieldand determine
exact importancein any particularcase. theactual spreadof certainmarkergenes.
Since the firstpublicationson the subject, Until such data are available, any gen-
the phenomenonof introgressive hybridi- eralizationsare based on mere opinions.
zation has been confirmed,with experi- Introgressionthereforegives us a par-
mentalverification, in numerousgenera. tial answerto our originalquestionas to
(A comprehensive bibliographyon intro- why hybridsare so seldom met with in
gression by Dr. Charles Heiser is now nature. It is because when hybridsdo
under way. The followingpapers are occur,theyusuallyperpetuatethemselves,
representative:Goodwin, 1937b; Riley, ifat all, in backcrossesto one or the other
1938, 1939a, 1939b; Dansereau, 1941; parentalspecies and the mongrelnature
Marsden, Jones and Turrill, 1946; Ep- of their descendantsis not apparent to
ling, 1947; Stebbins, Matzke, and the ordinarybiologist. The commonest
Epling,1947; and Heiser, in press.) Cir- result of hybridizationis introgression,
cumstantialevidenceindicatesits impor- and introgressionmust be excessive be-
tance in many more genera. How im- fore it will produce results conspicuous
portantis it on the whole in the higher enoughto impressbiologistswho are not
plantsand in othergroupsof organisms? makinga deliberatesearch for such phe-
We do not yet have any exact evidence. nomena. This is only half an answer.
From those cases where experimenters Why do interfertile species limit them-
have gone to the troubleof makingex- selves very largelyto backcrosseswhen
perimentalbackcrosses (Epling, 1947), they meet under natural conditionsand
it is clear thatmanyof the second back- mostparticularly whyis the backcrossing
4 EDGAR ANDERSON

largelyin areas wherenaturalconditions studiedthe secondgenerationfroma spe-

have been very much disturbed? There cies cross has noted the segregationand
are at least two main reasons; one re- recombinationof such physiologicaldif-
sides in the germ plasm itself,the other ferencesas length of blooming season,
in the habitat. As to the internalone, resistanceto diseases, daybloominghabit
the total effectof all the forces which vs. night blooming,ease of wilting,re-
make for specificcohesionis very great, sistanceto cold, lighttolerance,etc. If,
muchgreaterthanone would expectuntil therefore, we cross two species differing
he made carefulcalculations(Anderson, in their ecological requirementswe may
1939a & b). Following the arguments expect these physiologicaldifferencesto
used in these calculationsit can further- segregate as follows: The first hybrid
more be shownthatin well-differentiatedgenerationwill be uniformin its require-
species the total effectof linkage is so mentsand on the whole theywill be for
strong that two well-differentiated but conditionsintermediate betweenthose re-
interfertile species,meetingin an idealized quired for the two parents. The second
environmentfavorable to hybridization, generationwill be made up of individuals
would remainrecognizableunits in spite each of whichwill requireits own pecu-
of theirinterfertility. The details of the liar habitat.
argumientare largely mathematicaland Let us repeatthis last statement;it is
are shortlyto appear elsewhere,but the the crux of the argument. THE SECOND
general conclusionscan be terselyput. GENERATION WILL BE MADE UP OF INDI-
Linkage by itselfis a forcestrongenough VIDUALS EACH OF WHICH WILL REQUIRE
to prevent the complete swamping of ITS OWN PECULIAR HABITAT FOR OPTI-
interfertile species. As a factorin spe- MUM DEVELOPMENT. As a whole the re-
cific cohesion it is proportionalto the quirementsof the second generationwill
differentiation of the two hybridizing en- rangefroma need forsomethingmoreor
tities: the greaterthe differentiation, the less like one parental habitat to some-
strongerthe cohesiveforceof linkage. thingmore or less like the intermediate
The effectof the habitat,however,is habitatof the F-1 to somethingmore or
also important, and it usuallyoperatesin less like the habitatof the other parent.
exactly the same direction. The argu- In nature thereforewe mightreason-
mentis as follows: it is now knownthat ably expect to find firstgenerationhy-
the physiologicaldifferences betweenspe- brids growing in an intermediatezone
cies segregate in the same way as do betweenthe two parental
habitats. The
the morphologicalones. In Neurospora
persistenceof any considerablevarietyof
(Beadle, 1945) the mode of inheritance
the various secondgenerationrecombina-
of scores of physiologicaldifferences are
preciselyunderstood. In yeaststhe Lin- tions would requirea habitatsuch as is
degrens (1947) have demonstratedwith seldomor nevermet with,wherevarious
laboratoryprecision the inheritanceof combinationsof the two parentalhabitats
various differences in habitatpreference. are found in close juxtapositionto one
The higherplantsare not so amenableto another.
precisephysiologicalanalysesof theirnu- As a crudeexamplelet us considerthe
tritionalrequirementsbut thereis abun- adjacent habitats in which one finds
dant circumstantial evidencethat a simi- Tradescantiasubasperaand Tradescantia
lar situationprevailsand thereare precise canaliculataat homein the Ozark Plateau
data for a few characterssuch as ma- (Anderson and Hubricht, 1938). The
turity(see for instanceGoodwin 1937c, formergrows in deep rich woods at the
footnote13, or Marsden Jonesand Tur- foot of bluffswhile the lattergrows up
rill, 1946), response to day length,etc. above in fullsun at the edge of the cliffs.
Nearly everyone who has grown and As an over-simplified example we can
 HYBRIDIZATION OF THE HABITA'T' 5

list three of the outstandingdifferences naturalarea beforesuch a set of variedly

betweenthese two habitatsas follows: intermediate habitatscould be provided?
It has beenverygenerallyrecognizedthat
richloam........... rockysoil
deep shade........... fullsun if hybridsare to survive we must have
leaf mouldcover...... no leaf mouldcover intermediatehabitats for them. It has
not been emphasized,however, that if
Tradescantia canaliculata and T. sub- anythingbeyondthe firsthybridgenera-
aspera are well-differentiated species; tion is to pull through,we must have
neitherone of themis by any means the habitats that are not only intermediate
closest relative of the other, yet Mr. but which presentrecombinations of the
Hubricht and I have found by actual contrastingdifferencesof the original
experimentthat not only can they be habitats. If the two species differin
crossed readily by artificialmeans but theirresponseto light,soil, and moisture
they do cross abundantlywhen left to (and what related species do not?) we
themselves in an experimentalgarden musthave varied recombinations of light,
(Hubricht and Anderson,1941). Even soil, and moistureto grow their hybrid
thoughbothhe and I were familiarwith descendants. Only by a hybridization of
the appearanceof these artificialhybrids the habitat can the hybrid recombina-
and thoughwe searchedforthemat many tions be preservedin nature.
points where the species were growing The actual inherentdifferences in eco-
verynear one another,we foundveryfew logical preferencewill of course be mnuch
of thefirstgenerationhybrids. The habi- more diverse than in the crude example
tats of the two species are strikingly dif- given above. The numberof different
ferent;in the Ozarks one seldom finds kinds of habitatsrequiredby the hybrids
the intermediate habitatin whichthe hy- will rise exponentially with the number
brid is able to germinate- and survive. of basic differences betweenthe species.
This is a more or less intermediatecon- With ten such differences, arounda thou-
dition,a gravellysoil, partial shade with sand different kinds of habitatwould be
some brightsunlight,and a light cover- needed to permitthe various recombina-
ing of leafmould. Imagine,however,the tions to finda niche somewhereas well
habitatwhichmustbe providedif we are suited to them as the original adjacent
to see the second generationrecombina- habitatswere to the two parentalspecies.
tions which we obtain in the breeding With only twentysuch basic differences
plot. If we consideronly the threecon- (and this seems like a conservativefig-
trastingcharactersof the habitat which ure) over a milliondifferent recombined
have been mentionedabove, our recom- habitatswould be needed. Under natural
binationswould requirethe followingsix conditionsanythinglike such a situation
new habitatsin additionto the parental is close to impossible. Ordinarilyit is
ones (these six representonly the ex- onlythroughthe intervention of man that
tremerecombinations;a whole series of it is even remotelyapproached. Even in
intermediates will also be required) thesecases thenew "RecombinationHab-
rich loam rockysoil itats" will largelybe limitedto habitats
full sun deep shade prettymuch like those required by one
no leaf mould leaf mould of the parental species, but which in a
rich loam rockysoil few characteristics approach the require-
full sun full sun mentsof the otherparent. We may ex-
leaf mould leaf mould pect that even in such disturbedhabitats
rich loam rockysbil therewill be back-uprecombinations not
deep shade deep shade greatlydifferent fromone of the parents
no leaf mould no leaf mould
whichwill mostreadilyfindan ecological
What would have to happen to any niche suitedto them.
6 EDGAR ANDERSON

One of the best demonstrationsyet izationupon the intervention of man has

publishedof the way in which man can been describedby a numberof authors
providestrangenew nichesof hybridre- (Darrow and Camp,1 1945; Anderson
combinationsis thatgivenby H. P. Riley and Hubricht,1938). It was discussed
(1938, 1939a) in his analysis of the hy- in some detailby Wiegand (1935) in his
bridizationbetween Iris fulva and Iris paper on "A naturalist'sexperiencewith
giganticaerulea,two species which differ hybridsin thewild." Marie-Victorinhas
strikinglyin theircolor,morphology, and given a vivid description(1922, p. 32;
ecologicaladaptations(Viosca, 1935). In 1935, p. 65) of its operationwhen the
one of the localitieswhichhe studiedin originalfloraof the St. Lawrence valley
detailon the Mississippidelta,a series of was largely replaced by fields and pas-
long, narrow farms run straightback tures. Epling, Stebbins,Dansereau, and
from the highway side by side, in the theirstudentshave commentedupon the
fashionset by the French settlers,with connectionin a numberof different gen-
almosttheprecisionof experimental plots. era. Does this mean that introgression
The original environmentat that point as a phenomenonis limitedto the areas
was fairly uniform,but each man has disturbedby man and that its resultsare
treatedhis farm a little differently.It mere artifactsand not genuine natural
was strikinglyapparent from Riley's phenomena? I thinknot. Though freely
studythat the numbersand kinds of hy- admittingthatnearlyall the introgression
brids varied fromfarmto farm. Some which has been studied experimentally
had fewor none,whileothers,even when (for one exceptionsee Dansereau, 1941)
adjacent,had hybridsin greatquantities;' is of the nature of an artifact,I believe
there were significantly more of them thatat particulartimes,and in particular
where the meadows had been pastured. places, introgressionmay have been a
In one farm in particular,the little de- general evolutionaryfactor of real im-
pressionwhich ran parallel to the high- portance.
way had been subjected to a series of Under the conditionsof an experimen-
operations. The treesand shrubshad not tal garden, natural selectionamong the
been entirelyremovedfromthisarea, but progeny of a cross between species is
it had been repeatedlycut over and had much less severe than it is in nature.
in additionbeen heavilypastured. It had Though the optimumenvironmentsfor
a swarm of different hybridderivatives, the sisterhybridsmay be quite various,
almost like an experimentalgarden,and it is possibleto raise themajorityof them
the hybrid area went right up to the in one plot,providingthattheyare widely
fencelineat the border of the farmand spaced and competitionwith aggressive
stoppedthere. weeds is keptat a minimum. The preva-
Nor is this an isolated instance. lence of iris hybridson one or two of the
Viosca (1935) and other students of farms described by Riley (1938) may
Louisiana iriseshave workedout in con- have been due in part to the reduction
siderable detail the relationbetweenthe of competitionwith ot-herplants, par-
productionof hybrid swarms of these ticularlygrasses,as well as to the varia-
conspicuously different -irises and the tionsin shade and moisturebroughtabout
churningand rechurningof the habitat by repeatedrecuttingsand overpasturing.
by ditching,pasturing,lumbering,road- There must have been various times,
building,etc. It is only where man has even without the interventionof man,
,hybridizedthe natural environmentsof
1 Darrow and Camp also considered the re-
the Mississippi leltathatnaturecan find
action of hybridizingpolyploid complexes with
an appropriatelodgingplace for the hy- the environment. Polyploidy introducesfurther
brids she has created. complications into hybridizationwhich are be-
This dependenceof interspecifichybrid- yond the scope of this paper.
 HYBRIDIZATION OF THE HABITAT 7

when species hybridizedunderconditions tias fromMexico would have metspecies

whichproducedvaried new habitatsand coming down fromthe Appalachians in
when competitionwas not too keen, as an area conduciveto the survivalof some
forinstancewhennewlycolonizableareas of the hybridrecombinations.
emergedfromthe sea or when various Woodson has recently(1947) called
florasspread out onto the northernlands attentionto the importanceof peninsular
denuded by Pleistocene glaciation. At Florida in the 'speciationpatternsof the
such timesintrogression would have been easternUnited States. During parts-of
an importantevolutionaryfactor. For the Tertiaryit was an island or group of
one area we are beginningto get actual islands which finallybecame attachedto
proofthat it did occur. Along the coast themainland. Speciesand varietieswhich
of Californiathere are peninsulaswhich became differentiated during the island
once were isolated islands but whichare period must then have had unusual op-
now unitedwiththe main land. In their portunitiesto hybridizewith their rela-
studiesof the Californiaknobconepines, tives on the mainland. Giles' (1942)
fossiland living,Mason and his students studies of Cuthbertiain this area have
are demonstrating the actual role of in- given cytologicalproof that such hy-
trogression(for a general summarysee bridizationdid actuallytake place. Care-
-Cain,1944,pp. 112-118) in forming th,se ful studies of variation throughoutthe
pines as we know them today and to whole area, fora series of species,should
determinein some detail how introgres- yield data with which we could assess
sion operated at the time when these the general overall importancethere of
islands were joined to the coast. introgiression.
The Edwards Plateau in centralTexas
SUMMARY
is another area in which introgression
may have operated on a grand scale. 1. Experimental evidence shows that
This comparatively small area is a center will
sterility not account forthe rarityof
of distributionand variationfor numer- hybridsunder naturalconditions.
ous genera. A mereleafingthroughof a 2. Careful field analyses have shown
series of monographsof NorthAmerican that naturalhybridization is largelylim-
genera (Larsen, 1933; Anderson and ited to backcrosseswhich resemblethe
Woodson, 1935; Barkley,1937) willdem- parental species so closely that special
onstratethat it is one of the outstanding methods are required to detect them
centerseast of the Rockies. For many readily.
genera the concentrationof species is 3. One of the factorslimitinghybridi-
highertherethan at any otherpoint and zation to such introgressionis imposed,
for the genus Tradescantiawe know the by the habitatfor the followingreason:
even moresignificant factthatit is a cen- Two species differing in theirhabitatre-
ter for the diploid strains of polyploid quirementswill producea firstgeneration
species (Anderson and Sax, 1936; An- hybrid adjusted to a uniforminterme-
derson, 1937). The geological evidence diate environment.The second genera-
shows that when the Edwards Plateau tion howeverconsistsof individualseach
came into being, it united older land of whichrequiresits own peculiarhabitat
masses in Mexico and in the United for optimumdevelopment. Such hetero-
States. Certainlyat such a time related geneoushabitatsare seldomor nevermet
species in many genera mighthave met with, the only approach to them being
and hvbridizedunder conditionswhere found in places where man has greatly
competitionwould not have been keen altered naturalconditions.
and where associationsof plants were in 4. It is concludedthat hybridswarms
the makinginsteadof alreadyexistingas can surviveonlyin "hybridizedhabitats."
tightlyclosed corporations. Tradescan- While most of the latterresultfromhu-
8 EDGAR ANDERSON

man intervention,similarconditionshave DANSEREAU, PIERRE. 1941. Etudes sur les

prevailedin pre-humantimes when new hybrides de Cistes. VI. Introgression dans
la section Ladanium. Can. Jour. Research,
lands were opened up to colonizationby 19: 59-67.
diversefloras. At such timesand places DARROW, GEORGE M., AND W. H. CAMP. 1945.
introgressivehybridizationmust have Vaccinium hybrids and the development of
played an importantrole in evolution. new horticulturalmaterial. Bull. Torr. Bot.
Club, 72: 1-21.
LITERATURE CITED EAST, E. M. 1913. Inheritance of flower size
in crosses between species of Nicotiana. Bot.
A7NDERSON, EDGAR. 1936a. A morphological
Gaz., 55: 177-188.
comparison of triploid and tetraploid inter- . 1916. Inheritance in crosses between
specific hybrids in Tradescantia. Genetics, Nicotiaiia Lantgsdorffiiand Nicotiana alata.
21: 61-65.
Gelletics, 1: 311-333.
. 1936b. An experimental study of hy- EPLING, CARL C. 1947. Natural hybridization
bridization in the genus Apocynum. Ann. of Salvia apiana and Salvia mellifera. Evo-
Mo. Bot. Gard., 23: 159-168. lution, 1: 69-78.
. 1936c. The species problem in Iris.
FOCKE, W. 0. 1881. Die Pflanzen-mischlinge.
Ann. Mo. Bot. Gard., 23: 457-509.
569 pp.
. 1936d. Hybridization in Axnerican
GILES, N. H., JR. 1942. Autopolyploidy and
Tradescantias. Ann. Mo. Bot. Gard., 23:
511-525. geographical distributionin Cuthbertia gra-
. 1937. Cytology in its relation to tax- miiineaSmall. Amer. Jour. Bot., 29: 637-
onomy. Bot. Rev., 3: 335-350. 645.
. 1939a. The hindrance to gene recom- GOODWIN, RICHARD H. 1937a. Notes on the
bination imposed by linkage: an estimate of distributionand hybridorigin of X Solidago
its total magnitude. Am. Nat., 73: 185-188. asperiula. Rhodora, 38: 22-28.
. 1939b. Recombinationin species crosses. . 1937b. The cytogeneticsof two species
Genetics, 24: 668-698. of Solidago and its bearing on their poly-
-- AND LESLIE HUBRICHT. 1938. The evi- morphyin nature. Am. Jour. Bot., 24: 425-
dence for introgressive hybridization. Am. 432.
Jour. Bot., 25: 396-402. 1937c. The role of auxin in leaf de-
AND EARL HORNBACK. 1946. A genetical velopment in Solidago species. Am. Jour.
analysis of pink daffodils: a preliminaryat- Bot., 24: 43-51.
tempt. Jour. Cal. Hort. Soc., 7: 334-344. HUBRICHT, LESLIE, AND EDGAR ANDERSON.
AND RUTH PECK OWNBEY. 1939. The 1941. Vicinism in Tradescantia. Am. Jour.
genetic coefficients of specific difference. Bot., 28: 957.
'Ann. Mo. Bot. Gard., 26: 325-348. LARSEN, ESTHER L. 1933. Astranthium and
AND KARL SAX. 1936. A cytological related genera. Ann. Mo. Bot. Gard., 20:
monograph of the American species of 23-44.
Tradescantia. Bot. Gaz., 97: 433-476. LINDEGREN, CARL C., AND GERTRUDE LINDEGREN.
AND BRENHILDA SCHAFER. 1931. Species 1947. Mendelian inheritance of genes af-
hybrids in Aquilegia. Ann. Bot., 45: 639- fecting vitamin-synthesizingin Saccharo-
646. myces. Ann. Mo. Bot. Gard., 34: 95-99.
. 1933. Vicinism in Aqilegia vulgaris. MARIE-VICTORIN, FR. 1935. Flore Laurentienne.
Am. Nat., 67: 1-3. Montreal. 917 pp.
AND W. B. TURRILL. 1938. Statistical . 1922. Esquisse systematique et eco-
studies on two populationsof Fraxinuits. New logique de la flore dendrologique. Contrib.
Phytol., 37: 160-172. Lab. Bot. de l'Univ. de Montreal, no. 1: 1-33.
AND R. E. WOODSON. 1935. The species
MARSDEN JONES, E. M., AND W. B. TURRILL.
of Tradescantia indigenous to the United
1946. Researches on Silene maritima and
States. Contrib. Arn. Arb., 9: 1-132.
S. viulgaris. Kew Bull., 1946: 97-107.
BARKLEY, F. A. 1937. A monographic study
of Rhuts. Ann. Mo. Bot. Gard., 24: 265-498. MASON, H. L. (see Cain, 1944).
BEADLE, G. W. 1945. Biochemical genetics. MAYR, ERNST. 1940. Speciation phenomena in
Chem. Rev., 37: 15-96. birds. Am. Nat., 74: 249-278.
CAIN, STANLEY A. 1944. Foundations of plant REHDER, ALFRED. 1940. Manual of cultivated
geography. Harpers. 1-556 pp. trees and shrubs. Second edition. 996 pp.
CALVERT, A. E. 1929. Daffodil growing. Macmillan.
I
Dulau. 1-389 pp. RILEY, H. P. 1938. A character analysis of
CLAUSEN, JENS, D. D. IKECK, AND W. M. colonies of Iris fidva, Iris hexagorna var.
HIESEY. 1946. Experimental taxonomy. giganticaerulea, and natural hybrids. Am.
Carn. Year Book, 45: 111-120. Jour. Bot. 25: 727-738.
 HYBRIDIZATION OF THE HABITAT 9

. 1939a. The problem of species in the VIOSCA, P. 1935. The irises of southeastern
Louisiana irises. Bull. Am. Iris Soc., 3-7. Louisiana. Bull. Am. Iris Soc., 57: 3-56.
. 1939b. Introgressive hybridizationin a WIEGAND, K. 1935. A taxonomist's experience
natural population of Tradescantia. Genetics, with hybrids in the wild. Science, 81: 161-
24: 753-769. 166.
ROBERTS,H. F. 1929. Plant hybridizationbe- WOODSON,R. E., JR. 1947. Notes on the his-
fore Mendel. 374 pp. Princeton. torical factor in plant geography. Contrib.
STEBBINS, G. L., E. G. MATZKE, AND C. EPLING. Gray Herbarium, 165: 12-25.
1947. Hybridization in a population of ZIRKLE, CONWAY. 1935. The beginnings of
Quercus imarilandica and Quercus ilicifolia. plant hybridization. 231 pp. Univ. of
Evolution, 1: 79-88. Pennsylvania, Phila.