c

GPO PRICE S 150 NASA SP-100

I
I 2FSTI PRICE(S) S

t Hard copy (HC)

I Microfiche (MF) A

Solar Physics
1958-1964
I

I N66 2 5 0 0 3
I
1
.

Solar Physics
1958-1964

Scientific and Technical Inforiiiatron Dictsion

1966
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION
Washington,D.C.
FOR SALE BY T H E SUPERINTENDENT O F DOCUMENTS, U.S. GOVERNMENT PRINTING OFFICE
WASHINGTON, D.C.. 20402 ~ PRICE 50 CENTS
Foreword

T HIS VOLUME IS ONE OF A SERIES which summarize the

progress made during the period 1958 through 1964
in discipline areas covered by the Space Science and
Applications Program of the United States. In this way,
the contribution made by the National Aeronautics and
Space Administration is highlighted against the back-
ground of overall progress in each discipline. Succeed-
ing issues will document the results from later years.
The initial issue of this series appears in 10 volumes
(NASA Special Publications 91 to 100) which describe
the achievements in the following areas: Astronomy,
Bioscience, Communications and Navigation, Geodesy,
Ionospheres and Radio Physics, Meteorology, Particles
and Fields, Planetary Atmospheres, Planetology, and
Solar Physics.
Although we do not here attempt to name those who
have contributed to our program during these first 6 years,
both in the experimental and theoretical research and in
the analysis, compilation, and reporting of results, never-
theless we wish to acknowledge all the contributions to
a very fruitful program in which this country may take
justifiable pride.
HOMER E. NEWELL
Associate Administrator for
Space Science and Applications, NASA
 Preface

I
I T HE SUN IS A NATURAL OBJECT for the application of space
technology. The importance of mapping its spectrum in the
ultraviolet has long been recognized in establishing the presence
~ of atoms not represented by spectral lines in the visible region, or
1 in confirming tentative identifications based on faint subordinate
lines in the visible. The response of the Earth’s ionosphere to
solar flare radiation raised a question of the nature of that ra-
diation which could be directly answered by comparative ob-
I servations of Lyman-alpha and X-radiation during flares and at
quiet times. The emission lines of the corona, which are ob-
served as forbidden transitions in the visible region, predicate
permitted transitions of vastly increased strength in the extreme
ultraviolet. The detection and measurement of these lines
would permit a direct test of theories of excitation equilibrium
processes in the corona. The spectral energy distribution in the
~ Sun’s extreme ultraviolet spectrum has to be known in order to
I
~
interpret the state of excitation, ionization, and dissociation of
atoms and molecules in the Earth’s upper atmosphere. These
are among the most familiar examples of problems of solar phys-
ics which attracted the attention of space scientists almost as soon
as rocket technology became available for research in this coun-
try. Moreover, the highly variable and unpredictable charac-
I ter of solar activity, as, for example, solar flares, the corpuscular
streams which produce sudden commencement distrurbances in
the Earth’s magnetosphere, and the coronal enhancements asso-
I
ciated with the development of solar active regions, all created
the need for continuity of observations of the Sun above the at-
I mosphere, which only satellites could provide.
The study of the Sun from space vehicles must be viewed at
all times in relation to the entire discipline of solar physics, which
~ V
 b

PREPACE

utilizes a wide variety of distinct techniques of investigation.

The theoretician utilizing large digital computers to study the
constitution and evolution of the Sun, the radioastronomer who
studies structures in the upper levels of the corona by observing
occultations of radio stars, the laboratory spectroscopist who in-
terprets the spectra of highly ionized light elements, and the ama-
teur astronomer who counts sunspots as part of the international
observing program, all these specialists contribute in their par-
ticular way to the general goals of solar physics. Particular in-
terest in the Sun as the nearest and most readily observable star
has resulted in a specialization of astronomical research since
Galileo first counted and plotted sunspots. The peculiarities of
instrumentation necessary to exploit the brilliance and propin-
quity of the Sun sometimes lead us to forget that solar physics
is just a part of a broader field-stellar astrophysics. Similarly,
space solar physics will at times be discussed apart from the whole
discipline of diverse specialized techniques.
Astronomy differs from some branches of physics in the way
we go about obtaining new knowledge and, indeed, in the kind of
new knowledge we seek. For the Sun, particularly, the phe-
nomena are so complex that a purely qualitative description of
a phenomenon may be the closest we can get to a true understand-
ing of the phenomenon. In this respect the study of solar ac-
tivity is like the study of terrestrial weather. The parallel has
a further application-both sciences deal with unpredictable
events, and so require continuous monitoring observations to se-
cure needed data for research. The ideal of some sciences, to
relate complicated observations to a simple physical theory (for
example, by demonstrating that the observed boundary condi-
tions are compatible with a simple differential equation) , often
cannot be achieved in solar astronomy. Some of the most grati-
fying advances made in solar physics have indeed been the discov-
ery that complicated things have simple explanations: for ex-
ample, the identification of the coronal emission lines, and the
interpretation of sunspot motions as the migration of magnetic
current systems from high to low latitudes during a double solar
cycle. But all the outstanding problems of solar physics, insofar
as they have been resolved, have merely raised other more diffi-
vi
the specialist. Progress has to be made in orderly steps of in-
creasing resolution, corresponding in a rough way to the scale of
the solar structures of interest. These range from active regions
up to a few minutes of arc in size; through flares and sunspots,

down to the smallest structures we know, including the fiber

structures of prominences, chromospheric spicules, and the small-
 Contents

I cbapter Page
1 BACKGROUND ...................................... 1
2 THE FRONTIERS OF SOLAR RESEARCH ............ 5
3 INSTRUMENTATION.. ..................... ..... ...
. 13
4 MAPPING THE SUN’S ULTRAVIOLET SPECTRUM. 33
5 MAPPING THE S U N S SOFT X-RAY SPECTRUM.. 49
6 EVIDENCE OF SOLAR HARD X-RAY EMISSION.. . 57
7 CORONA ............................................ 59
8 SATELLITE MONITORING OF SOLAR RADIATION. 63
9 THE FUTURE FLIGHT PROGRAM .................. 81
10 SUMMARY AND CONCLUSIONS ................... 87
..........................................
REFEENCES 91
 1

Chapter 1

' Backgrozlnd

A POWERFUL NEW TOOL became available to the astronomer in

the form of V-2 rockets, and its most obvious application
I was to solar studies. The first objective was to chart the Sun's
ultraviolet spectrum below the atmospheric window cutoff at
about 2910 angstroms. Mapping of the solar ultraviolet spec-
' trum has continued up to the present and probably will never be
complete as long as wavelength resolution or the combination of
wavelength and angular resolution can be improved. A second
I
type of rocket-borne experiment measured the spectral distribu-
tion of solar radiation. Interest in the Sun's far ultraviolet has
been stimulated by theoretical models of the Earth's upper at-
mosphere and ionosphere. Such models must assume an intense
and highly variable source of solar ionizing radiation. Either
hydrogen Lyman-alpha, the resonance radiation of the most
~
abundant element in the Sun, or X-rays had to be invoked to
account for some of the principal characteristics and variations
~

of the ionosphere. The identity of this ionizing radiation re-

mained an object of speculation and debate until rocket obser-
vations finally settled the question during the International
Geophysical Year (IGY). '

'
Since these initial explorations of the Sun's far-ultraviolet and
X-ray spectra, our knowledge has increased considerably. The
purpose of this review is to cite the advances in solar astronomy
made possible since the IGY by space research techniques. This
time period overlaps the history of the National Aeronautics and
Space Administration. However, the review will not be con-
fined to NASA efforts. Much of the research reviewed was per-
formed by laboratories of other agencies, notably the Air Force
Cambridge Research Laboratory (AFCRL ) and the Naval Re-
search Laboratory (NRL). Part of this space research by other
1
 b

SOLAR PHYSICS

Government agencies was of course, sponsored and supported by

NASA.
To set the stage for this review, it may be helpful to highlight
the progress in solar research by rockets and satellites as reported
at the General Assemblies of the International Astronomical
Union in 1958. At the beginning of solar cycle 19 it was thought
that the X-ray spectrum of the quiet Sun could be approximated
by a half-million-degree blackbody distribution. This tempera-
ture implies that the X-ray spectrum has an effective short wave
limit near 20 A. When coronal condensations at a temperature
of 2 million degrees were present however, the X-ray spectrum
extended down to as low as 6 A. During the preceding solar
cycle (no. 18) the highest and lowest values of X-ray flux from
the quiet Sun were 0.15 and 1 erg/cm'/sec, respectively. In the
spectral region between 1050 8, and 1350 A, 90 percent of the
total intensity results from hydrogen Lyman-alpha radiation.
During solar cycle 18, fluxes ranged from 0.1 to 1.6 ergs/cm'/sec.
By 1958 intensities as high as 6 ergs/cm"/sec had been detected.
The NRL rocket program during the IGY had obtained flux
measurements during flares of importance 1, 2, and 3, showing
that X-ray emission below 8 8, was greatly increased, and prob-
ably extended down to 1 8, or 2 A. Gradients in the continuous
spectrum had been measured spectrophotometrically to wave-
lengths below 2200 8, and were found to correspond to an
effective temperature of 5200".
The Fraunhofer spectrum, which predominates down to about
2085 8,) had been mapped extensively. The Mg I1 doublet
(2795 8, and 2803 8,) was recognized as the most important fea-
ture of this spectral region. A rocket-borne echelle grating spec-
trograph provided an effective resolution of 0.1 A down to 2000
A. The doubly reversed Mg I1 lines were shown to be very sim-
ilar to the H and K lines of Ca 11, except that the H? and Kz
components were stronger in Mg 11.
Below 2085 8, the continuum intensity suddenly decreases, and
the Fraunhofer lines become less distinct. T o account for this,
it was suggested that molecules such as NO and CO in the Sun's
2
 BACKGROUND

atmosphere were acting as quasi-continuousabsorbers. Between

2000 A and 1500 A, the solar spectrum was known to change
from a continuum with Fraunhofer lines to a region where emis-
sion lines emerged above a rapidly weakening continuum.
Rocket spectra in the range 1000to 2000 A, made in 1955 by both
NRL and the group at the University of Colorado, revealed
nearly all of the strong resonance lines of the most abundant ele-
ments in the Sun (hydrogen Lyman-alpha and Lyman-beta ; He
11; C I; C 11; C 111; C IV; N V ; 0 I a n d 0 VI; Si 11, Si 111, Si
IV; S I1 and P 11). Certain expected lines were absent, such as
the raies ultimes of N I, N 11, and N 111. Rense had made a
single spectrum extending to much shorter wavelengths, including
the resonance line of He I1 at 304 A, and the He I resonance line
a t 584 A.
The width of the hydrogen Lyman-alpha line was not known
with certainty. One observation suggested that the line is as
wide as 0.8 A, another that it is no greater than 0.3 A. Tenta-
tive hydrogen Lyman-alpha spectroheliograms were only partly
successful and barely sufficient to indicate the general conformity
between plages as shown by hydrogen Lyman-alpha, hydrogen
Balmer-alpha, and calcium K-lines. Spectrograms made with
a stigmatic spectrograph showed that limb brightening occurs in
certain lines (C IV) and not in others ( C I ) . The results were
interpreted in terms of a simple, stratified, semitransparent
atmosphere.
In short, much was known about the Sun’s ultraviolet and
X-ray spectrum by 1958. This report summarizes the progress
made from 1958 to 1964.

3
 chapter 2

The Frontiers of Solar Research

T HE PRINCIPAL TASK OF SOLAR ASTRONOMY is to analyze the

many kinds of electromagnetic radiation emitted by the
Sun. This radiation must be observed at all wavelengths,
including those to which the Earth's atmosphere is opaque.
Thus solar telescopes, spectrographs, and other instruments must
be carried above the atmosphere in rockets, satellites,and probes.
Some studies require continuous observations over days or
months, as in studies of solar activity. These must ultimately
use equipment carried in satellites which are independent of
terrestrial meteorology and diurnal effects. Real-time moni-
toring of solar events will be supported by such satellites, which
will telemeter data to ground stations.
Angular resolution of solar surface details is limited from the
ground by atmospheric turbulence. By placing telescopes above
the atmosphere, the theoretical limit of resolution for a perfect
lens or mirror can be attained.
Another reason for observing the Sun from rockets and satel-
lites is atmospheric scattering. The inhomogeneities and dust
of the atmosphere scatter the Sun's disk light, making it impos-
sible to see the outer solar corona, the zodiacal light, or the
gegenschein free of contamination. Space platforms, being
above the scattering layers, are not subject to this disadvantage.
Radiofrequency observations of any celestial source below
about 10 Mc/sec must be conducted above the interfering layers
of the Earth's ionosphere. The electrically conducting plasma
in the upper atmosphere absorbs completely the solar radio
emission below a critical frequency which diminishes with height.
Even above the critical frequency for ground observations, large-
scale irregularities and time variations in the ionosphere produce
scintillation (fluctuation) in solar radio emission observed from

5
 SOLAR PHYSICS

the ground, which masks solar phenomena. Low-frequency

ground observations of cosmic radio noise are much confused
by terrestrial transmissions, which are scattered and reflected
downward by the ionosphere, so that an orbiting low-frequency
solar radio telescope will also benefit from diminished back-
ground noise.
T o describe the present state of knowledge and the major areas
of interest, some of the frontiers of solar research will be dealt
with from the standpoint of the space astronomer as well as his
colleagues who view from the ground.
SOLAR ACTIVITY

Neither the origin nor the effects of the time variation of solar
processes are well understood. Nearly every true statement
about solar activity can be properly rephrased as a question. We
know that solar activity varies over an 11-year period but can-
not show how this period results from boundary conditions.
One goal of the solar astronomer is to discover the relation of
the Sun’s basic parameters (diameter, luminosity, composition,
age, and rotation) to the time and energy scales of solar activity.
Another unexplained phenomenon is the familiar migration of
centers of activity from middle latitudes toward the Equator
during the 11-year cycle. The outstanding tasks of this discipline
include solving the following problems :
( 1) What causes solar flares?
( 2 ) How are solar protons accelerated and expelled?
( 3 ) What causes the persistent plasma emission (the M-re-
gion source) from activity centers?
SOLAR MAGNETISM

Although solar magnetism is known to be closely linked to solar

activity, the relationship is poorly understood. Sunspots exhibit
the strongest magnetic polarities on the surface. They seldom
appear as simple dipoles or as the result of simple electric current
systems. The elementary theory of magnetic diffusion suggests
a natural sunspot lifetime of a few hundred years, in contrast to
their observed lifetimes of a few weeks. Theories of the origin
and dispersal of sunspot magnetic fields have to take into account
6
 THE FRONTIERS OP SOLAR RESEARCH

the resistance of the photospheric plasma to changes in magnetic

fiilds. The usual explanations of rapid changes involve bulk
transport of magnetized material by hydromagnetic processes.
The microscopic relation of solar magnetism to velocity and ther-
mal structure remains one of the most pressing observational
tasks of solar astronomy. The scale of such correlative observa-
tions is limited by the resolution of ground telescopes to about
one-tenth the otherwise attainable Doppler or direct resolution.
Important questions of immediate interest in solar magnetic
observations concern magnetic properties of structures in the
whole size range from a few hundred kilometers (prominence
fibers, granules) through a few thousand kilometers (sunspots,
prominence gross structure) to a few hundred thousand kilo-
meters (global distribution of magnetic polarity and field
strength). Many specific questions concerning solar magnetism
remain unanswered, including the cause of north-south hemi-
spheric spot polarity alternation and the alternation of leading-
spot polarity between successive sunspot cycles. A theory of
spot magnetism must account for the observation that some spots
are bipolar, others predominantly unipolar, and others hetero-
geneous. Observations, rather than theory, will probably reveal
how solar magnetism on the largest scale (general field) evolves
in relation to the 11-year solar cycle. Related to this phenome-
non is the enigmatic discovery that the magnetic fields at the
Sun’s N and S poles do not change polarity in phase with each
other.
At the other end of the known size scale, it is not yet determined
whether granules or chromospheric spicules are discretely mag-
netized. The relationship of their magnetic and geometric
properties to their kinematic behavior is the cornerstone of any
theory of these phenomena and probably of gross forms of solar
activity as well.
Ultimately, we seek to learn the origin of the Sun’s magnetic
fields. Solar rotation and the convective transport of subphoto-
spheric plasma have been suggested as the basic mechanism, but
these theories cannot yet predict the observed character of mag-
netic fields in spots or the general field. In this general area of

ZOD-S160--88---2 7
 SOLAR PHYSICS

research, exploratory observations will probably precede ad-

vances in theory for some time to come.
ATMOSPHERIC INHOMOGENEITIES AND TURBULENCE

At the limit of attainable ground resolution, photospheric

granulation reveals the vital role played by turbulent convective
energy transport in the steady-state configuration of the Sun.
Higher in the atmosphere off-band hydrogen spectroheliograms
disclose the vertical transition from microscopic turbulent energy
transport to a larger scale process. The supergranulation has
recently been discovered to be intermediate between these small
scale motions and the grosser inhomogeneities of sunspots and
activity centers. The relationship between the supergranula-
tion, solar magnetism, and active centers is only known in a crude
way and is not understood at all. All the most difficult tasks of
solar observation are united in this area, Weak-field magne-
tometry of small structures, monochromatic time lapse pictures
made simultaneously at several wavelengths, and velocity-dis-
criminating observations must be made continuously over ex-
tended periods of time.
In the higher layers of the solar atmosphere, different char-
acteristic inhomogeneities dominate. I n the chromospheric
transition layer, the spicules correspond to the small-scale turbu-
lent velocity field surrounding active regions. During the last
decade monochromators and high-resolution spectrographs were
first applied effectively to study spicules at the limb. Quite
recently, off-band techniques combined with time-lapse photog-
raphy have provided meaningful new observations of spicules
on the disk. The new data have not, however, answered the
basic question of why spicules occur.
The corona has been shown at many total solar eclipses to
have the shape of an oblate spheroid with discrete radial stream-
ers. The size and distribution of coronal features vary with the
solar cycle but also from day to day as the Sun rotates and as
the individual activity centers evolve. Indirect evidence of
inhomogeneities in the corona is provided by variations in the
temperature of the solar wind. These variations have been
observed directly by interplanetary space probes and inferred
8
 I

THE FRONTIERS OF SOLAR RESEARCH

.from their effect upon the Earth's magnetic field. The mecha-
nism of this influence is not understood.
More highly organized coronal f o r m are familiar as promi-
nences and are readily visible as condensations of partially ion-
ized gas having a variety of shapes. Most of the information
about these forms is puzzling: why they form, how they can
I persist, their relation to activity centers in the chromosphere,

I details of their fine structure, and their passive response to

chromosphericimpulsive activity.
The forms and variations in coronal structures at the l i b
can be observed by combining the coronagraph with the bire-
fringent filter. Satellites, however, will be required to study this
form of solar structure in sufficient depth to understand it.

'
I FLARE PHENOMENA

A flare is a localized, sudden brightening of the chromosphere

in H-alpha or the calcium K-line. Large flares often produce
associated radio noise bursts, sometimes cause sudden-
commencement magnetic storms, usually produce flashes of
X-rays, occasionally eject energetic protons, and (apparently
fortuitously) disturb prominences in the neighboring corona. A
vast amount of detail is known of these chromospheric brighten-
~
ings, but as yet it can only be guessed how a flare is formed;
why it occurs, where and when it does; or how it relates to the
concomitant nonthennal phenomena. The answers to this par-
ticular set of questions continue to rank among the principal
objectives of solar astronomers.
Numerous flare theories have been advanced during the past
few years, some of which could possibly be tested by direct
observation. Typical of these are theories postulating the trans-
I fer of magnetic energy to streams of charged particles by such
plasma mechanisms as pinch discharges and the Fermi process.
Observationsmade with unlimited wavelength, magnetographic,
angular, and time resolution over the whole electromagnetic
spectrum could conceivably test the validity of such theories.

I' One goal of solar astronomy is to test such theories by following

promising lines of specialized observations. A simple example
is the theory of magnetic relaxation of the entire field of a sun-
I
9
I
 SOLAR PHYSICS

spot group. This theory could be tested with magnetic maps,.

accurate to a few gauss and a few hundred kilometers, by com-
paring preflare with postflare sunspot configurations.
Familiar spectroscopic and cinematographic techniques will
continue for a long time to play a major role in flare studies, es-
pecially as advances in technology or superior observation sites
improve observational capability. For some questions, increas-
l
ing the resolution in the visible spectrum could be as important as
extending modest-resolution observations to an unexplored spec-
tral region. I n general, however, the greatest promise for un-
derstanding flares is offered by high angular resolution, and high
time resolution X-ray qnd ultraviolet studies of the rising phases
I
of flares, including the analysis of line profiles of critical emis-
sions. This observatiohal objective can be stated in terms of nu-
merous specialized observations, of which the following are only
typical : 5- and 1-arc-second resolution X-ray pictures of flares,
both in resonance lines (such as Lyman-alpha of C VI or 0
V I I I ) and in the continuum near 1 A and 10 A; maps of the
Sun’s face showing the variation in the profile of the hydrogen
I
Lyman-alpha line, especially the extended faint wings, with res-
olution of structures as small as 100 km; spectrograms in the en-
tire extreme ultraviolet region (100 A-3000 A ) of the hottest I

portions of flares, to obtain time variations of equivalent widths

of multiplets over the range of excitation and ionization energies; l
and polarization observations of the continuum radiation emitted
from flaring regions.
The radiofrequency emission of flares has been energetically
investigated during the past solar cycle. Together with direct
measurements of solar protons, this technique has revolutionized
our outlook on flares, on the Sun, and even on the universe.
Because wavelengths of a meter and more are involved, angular
resolution is only adequate to locate radio bursts relative to ac-
tivity centers. In theory, very large telescopes operating at cen-
timeter wavelengths could provide detailed radio-emission maps
of flares to show how their nonthermal and their quasi-thermal
processes are associated geometrically. The basic physical
mechanisms may be determined by polarization and spectral

10
 I T H E FRONTIERS OP SOLAR RESEARCH

I analysis at high angular resolution for investigating local mag-

netic fields or for discriminating between basic radiative proc-
esses. At very long wavelengths, low spectral resolution obser-
vations alone may provide additional information on the origin of
the radiofrequency emission.
RELATION BETWEEN GROUND AND SPACE OBSERVATIONS OF THE SUN
I

I From ground observatories we see principally the quasi-ther-

mal radiation of the Sun. The nonthermal emission, from which
the desired information on solar activity could be obtained, lies
near the threshold of detectability in the visible region. Only
with the most refined monochromators, Hale’s spectroheliograph
and Lyot’s birefringent filter, can this emission from the chromo-
sphere and corona be measured, because the Sun’s spectral lu-
minosity reaches peak intensity near 5000 A, right at the center
I of this atmospheric window. At the extremes of the electro-
magnetic spectrum, in the ultraviolet and radiofrequency re-
gions, the quasi-blackbody radiation is much weaker than the
nonthermal component.
Despite this handicap, observations from the ground must
play a dominant role in solar research for many years to come.
I Angular resolution of a high order is required for studying such
structures of solar activity as spots and flares and such inhomo-
geneities of the quiet Sun as spicules and granulation. The
~

atmosphere permits apertures up to about 30 inches to be used.

To achieve a resolution higher than about 0.3 arc-second, larger
aperture telescopes must be carried above the atmosphere; this
will be done by balloons andeventuallyby satellites. In the
meantime, solar observations from space vehicles serve best in
the inaccessible extreme spectral regions, using the modest
angular resolution possible by the size of telescope and preci-
sion of stabilization available. Radio emission from the Sun
offers a most powerful tool to discover and monitor nonequi-
librium processes from the ground. However, high-enough
angular resolution at centimeter, meter, and decameter wave-
1 lengths cannot be attained to investigate basic structures like
spicules, prominences, spots, and granulation, because at such

11
 SOLAR PHYSICS

wavelengths high-resolution telescopes grow to prohibitive.

dimensions.
In another way, satellite observations even today can, in prin-
ciple, excel ground observations. Continuity of observation is
essential to study adequately the rare and unpredictable solar
events of greatest interest, such as flares. Our understanding
of the supergranules is presently hampered by inadequate obser-
vations of their time development, as one observatory can follow
the evolution of an element only throughout part of its lifetime
during each day. Satellites carrying large telescopes in con-
tinuously sunlit orbits will eventually permit such studies to
proceed over arbitrarily long times without interruption by
sunsets and cloudy days.
Meantime, simple satellites permit uninterrupted monitoring
of the X-ray brightness of the Sun and could, if desired, provide
the equivalent of a chromospheric-flare patrol in hydrogen light.
That function is currently performed more economically from the
ground because no firm requirement exists for the guarantee of
complete monitoring. If a few gaps in coverage, up to a few
hours length can be tolerated, a relatively inexpensive H-alpha
patrol network can replace a more expensive orbiting monitor.
This situation may change as operations in space generate a firm
requirement for reliable, full-time monitoring.

12
 chapter 3
Instrzc mentation

' s IGNIFICANT PROGRESS has been made in developing the tools

of space solar astronomy. Since mere invention of a new
technique often enables a number of new scientific advances to
be made, we shall consider this technological progress in detail.
This topic has been reviewed comprehensively in reference 1.
SOLAR POINTING CONTROLS

Early rocket flights were made without any attitude-control

systems, and experimenters had to rely on the rocket's tumbling
to sweep their detectors across the desired field of view, together
with a Sun sensor to establish the detectors' attitude as a func-
I
tion of time. The NRL satellites SR I and SR I11 were also
unstabilized and their detectors operated in a sky-sweeping
mode.
In contrast, the present generation of satellites features many
sophisticated attitude control systems (ref. 2 ) . Before the in-
troduction of reliable 3-axis attitude control systems, satellites
such as the Explorer and Pioneer series successfully employed
spin stabilization to achieve single-axis stability. This technique
is still attractive, especially for equatorial or near-polar orbits
where uncontrolled spin stabilization can be maintained for
durations of 1 to 3 years.
Probably the most sophisticated active stabilization system to
date is that used on OS0 I, which is designed to point 75 pounds
of instruments at the Sun with an accuracy of 1 arc-minute. In
addition, 100 pounds of instruments are housed in a spinning
section of the spacecraft and scan the Sun once every 2 seconds.
Unlike passive systems, the spin rate on O S 0 I is maintained
within 5 percent of its nominal value by small reaction jets,
with a lifetime of at least 6 months. The spin axis itself is simi-
larly maintained within 3" (ref. 3).
13
 SOLAR PHYSICS

The Advanced Orbiting Solar Observatory ( AOSO ) , SUC-

c e w r to OSO, also will employ an accurate attitude contrbl
system which points the spacecraft at the Sun with an error of
less than 5 arc-seconds, a vast improvement on the systems men-
tioned earlier. The AOSO system can also operate in a scan
mode, producing a 60-line scan of a 5-arc-minute square posi-
tioned anywhere on command within a 40-arc-minute square
centered on the Sun. In addition, a coarse scan of the whole
40-arc-minute square is possible; this function is also performed
by the second and certain later OSO’s.
The effect of such accurate pointing systems on the scope and
design of solar experiments is not hard to see. While experi-
menters once had to be content with fleeting looks at some un-
specified part of the solar disk, they can now design experi-
ments to scan accurately specific areas of the Sun and observe
time-variant phenomena with the assurance that it is the phe-
nomenon that is changing and not the direction in which their
equipment is being pointed.

DETECTORS

Soft x-Rays (1 A-loo A)

The fact that most materials in the soft X-ray region are
highly opaque is a serious limiting factor in detection. Ioniza-
tion type detectors with extremely thin windows are commonly
used; photographic emulsions are also suitable. Scintillation
counters and devices using phosphors or semiconductors are
generally suitable for the hard X-ray, ultraviolet, and longer
wavelength regions ; however, they are inefficient in the soft
X-ray region because the X-rays are absorbed in the top few
microns of the detector surface. Photoelectric devices work sat-
isfactorily, although with lower efficiencies than in the ultra-
violet and visible regions. The X-rays penetrate the surface
sufficiently for some of the photoelectrons to be captured before
emission.
PhotograQhic Emulsions
The first detector used for solar soft X-rays was photographic
emulsion, with a thin shield of beryllium foil to mask visible
14
 INSTRUMENTATION

and ultraviolet radiation (ref. 4). Photographic emulsions

'have been used as detectors in pinhole cameras (ref. 5 ) , and
in photometers (refs. 6 and 7). In the soft X-ray region - a
linear relation exists between exposure and fil m density at low
film density, and emulsions have been used for absolute cali-
bration (ref. 8). The sensitivity of emulsions to X-rays is quite
high, approaching 1 developable grain per photon.
Pbotoionizatwn Detectors
Photoionization may be used to detect soft X-rays in ioniza-
tion chambers, Geiger counters or proportional counters. The
ionization counter is least sensitive and is used in very high flux
conditions (refs. 9, 10, and 11) and is commonly filled with
argon or nitrogen. Geiger counters will count single photons
and are simple and robust; they have been used for a number
of rocket and satellite observations (refs. 9 and 12).
The counter developed by NRL for solar soft X-ray work is
described by Newell (ref. 13). Proportional counters will also
work satisfactorily in the soft X-ray region up to about 20 A
and give moderate wavelength resolution by pulse-height anal-
ysis. Stable gas fillings and high voltages are required to ob-
tain constant and reproducible response. Proportional count-
ers have been developed at the University of Leicester (ref. 7)
and have been flown on Skylark rockets (ref. 14) and on Ariel I
(ref. 15). Typically the counters use a mixture of argon and
methane at a pressure slightly less than 1 atmosphere. The
methane acts as a quenching agent at high counting rates. Ion-
ization detectors are the most common in solar soft X-ray rocket
and satellite experiments conducted so far and will also be'in-
cluded on most of the O S 0 series of satellites. Typical spec-
tral responses are shown in figure 1.
The performance of ionization counters is considerably af-
fected by the absorption of the window. Only elements of low
atomic number can be present in the window because they have
the least absorption ; typical window materials are beryllium,
aluminum, mica, and plastic films such as Mylar, Glyptal, and
nitrocellulose. The absorption is strongly wavelength depend-
ent, especially if absorption edges are present in the soft X-ray
15
 SOLAR PHYSICS

region. Extremely thin windows are required, of the order of.

a few microns thick; or for wavelengths of 50 A, of less than 1
micron. This requirement presents some technical problems
which have been solved by using small-area windows, by SUP-
porting the thin film on wire mesh, and by using flow counters
to eliminate the effects of pinholes and leaks.
Window absorption usually determines the lmg wavelength
limit in photographic detectors and ionization counters; in some
cases it can be used intentionally to limit the response. The use
of multiple filters with an absorption edge in the required wave-
length range (as is the case with aluminum) enables a narrow
band to be isolated, but this technique is limited to a few bands
by the characteristics of available materials.

Photoelectric Detectors

In the soft X-ray region the quantum energy is high enough

to eject photoelectrons from most solid materials. As noted
earlier, however, the photoelectric efficiencies of common

‘ 90 0 ° 3

Wavelength, A

Figure 1 .-Typical spectral response of ionization counters. Window material and

thickness are indicated on the curves.

16
 INSTRUMENTATION

photocathode materials are generally lower than in the ultra-

violet or visible, but a number of unconventional cathode mate-
rials such as beryllium, tungsten, and titanium show an appreci-
able effect with soft X-rays and negligible emission in the visible
and ultraviolet (ref. 16). Detectors using these cathode ma-
terials, while having counting efficiencies l e s than Geiger or
proportional counters, can be used without windows in the space
I
environment. Photoelectric detectors for soft X-rays are
usually open-ended photomultipliers, and since a volume rather
than a surface photoelectric effect is involved, the cathode sur-
faces are less affected by exposure to the atmosphere than are
conventional materials. Photocathodes of Be0 and SrF,, have
been used in Russian detectors (ref. 17). Investigation of the
possibility of using transmission photocathodes to make soft X-
ray image converters and dissectors is included in the O S 0 and
AOSO programs.

Scintillation counters with beryllium or aluminum windows

and NaI-TI crystals have a response from 20 to 200 keV
( 10 keV= 1.25 A) , beyond the short wavelength limit of the
soft X-ray spectrum. With pulse-height analysis they can be
used to map the spectrum above 20 keV in a similar manner
tO the proportional counter.
Most soft X-ray detectors are also sensitive to the charged
particles in the radiation belts and in the solar wind. The low
energy particles are stopped by the windows if present, while
those of higher energy are usually deflected from the window
aperture by “broom” magnets. Even this arrangement may be
insufficient for particles of very high energy, and it is usual
in these cases to use either closed counters or special cosmic ray
detectors with anticoincidence techniques. In most cases, re-
gions of high particle flux can be avoided by suitable choice of
orbit. In the case of O S 0 I the energetic electron flux is avoided
by the nearly circular, 300-nautical-mile-high orbit, inclined at
30”. However, the regions of higher flux (e.g., in the South
Atlantic anomaly) are not completely avoided. Conversely, high
17
 SOLAR PHYSIC!?

inclination orbits will suffer most from transit of the trapped

radiation belt at high latitudes.
Ultraviolet
Photocells
For many years the only spectrally selective ultraviolet detec-
tors had been the pure metal type which resulted from the early
work of Rentschler et al. (ref. 18). Typical of these metals are
gold and nickel, with gold having better rejection at longer wave-
lengths. Another useful material is the composite surface
cesium-antimony,although the extension of its usefulness into the
vacuum ultraviolet has been slow. Over a period of some 15
years, however, it has been possible to extend the range of such
surfaces to nearly 1000A by the use of various windows. Dunkel-
man (ref. 19) describes this progress and has obtained a flat,
high efficiency ( > lo-' photoelectrons/quantum) down to 1200 A
for a Cs-Sb photomultiplier using a.CaF2 window. LiF windows
have extended the sensitivity to nearly 1050A.
A more significant advance in ultraviolet detection occurred
when several akali-tellurium surfaces were developed by Taft
and Apker (ref. 20). Since then numerous other workers have
prepared both opaque and transparent alkali telluride surfaces,
with some of the opaque surfaces having a quantum efficiency
near 50 percent at 2500 A. Notable among these are Cs-Te
and Rb-Te, which have nearly the same efficiencies and rejec-
tion ratios. The alkali tellurides have a high work function and
are particularly useful in applications where discrimination
against longer wavelengths is important. Other materials with
still higher work functions which are spectrally selective for the
vacuum ultraviolet region. are certain iodides and bromides ;these
substances are deposited on a thin conducting substrate of tung-
sten behind a LiF window. Typical for these detectors is a
peak yield between and unity near 1000 A and a cutoff
between 2300 A and 2500 A. In general, then, with suitable
windows, certain cathodes provide satisfactory yields to approxi-
mately 1000 A, with sharp cutoff or wider spectral response
available as desired.
18
 Ih'STRUMENTATION

.Another recent, significant development is the windowless

continuous-channel multiplier designed by Goodrich and Wiley
(ref. 21) which, because of its tiny size and high sensitivity, is
unique among photocells. This multiplier is simply a short glass
tube with its internal wall coated with a high-resistance layer.
A potential difference of 1000 to 2000 volts is maintained be-
tween the ends, resulting in a current on the inner surface and a
uniform axial electric field down the length of the tube. A
photoelectron ejected from the inner surface near the open end
is accelerated down the channel by the high voltage; multiplica-
tion occurs at each collision with the wall. The emitted elec-
trons cascade down the length of the tube, producing gains of
1Osor more.
A rather recent study of this multiplier (ref. 22) has shown it
to have a spectral response similar to that of tungsten, although
the photoemitting inner surface is an oxide of some other metal.
Thus it is perfectly suited to measurements in the extreme ultra-
violet, since it is insensitive to strong, near-ultraviolet and visible
radiation which appear as stray light in any spectrum produced
by a single grating.
Goodrich and Wiley (ref. 23) also developed another type of
tube which was modified in collaboration with Hinteregger for
the rocket-borne grazing incidence measurements. This is a
magnetically focused, strip-type photomultiplier, with multipli-
cation gain produced through crossed magnetic and electric
fields. Photoelectrons, after emission from the cathode, move
into the space between the two resistance strips where they follow
a trochoidal path down the corridor between the strips, produc-
ing many secondary electrons at each collision with the second-
ary emitting surface. A gain of lo*or more is achieved. This
type of phototube has been used successfully in the range of 55 A
to 1300A.
uyicon

The uvicon is an ultraviolet-sensitive image tube, the main

components of which are a photocathode, an electron imag-
ing section, an electron-bombardment-induced-conductivity
(EBIC) target, and a standard vidicon gun. The photocathode
19
 SOLAR PHYSICS

assembly may consist of a LiF lens, a semitransparent conductor

and a photoemitting material. The electrons released from the
photocathode are electrostatically accelerated and focused on the
EBIC target by a multielectrode system. The image size and
focusing may be controlled by lens-element voltage ratios.
Energetic electrons released from the photocathode strike the
target and generate a video signal. Each electron incident on
the target generates charge carriers whose number depends on
the voltage across the target material, the energy of the incident
electrons, the temperature of the target material, and the com-
position. One such tube, designed for astrophysical purposes
(ref. 24) had a gain of 500 for 40 volts on an arsenic trisulfide
target.
After this tube was evaluated for use in Project Celescope, cer-
tain changes were made until finally a transmission-secondary
emission target was employed. With this target, the minimum
detectable point image (i.e., the image for which peak signal
equals rms noise) contains about 1.4x lo3photoelectrons. This
value refers to measurements conducted at 30 frames/sec, 4
Mc/sec video bandwidth, with laboratory equipment. The
uvicon and related detectors have not yet been employed to ob-
serve the Sun. These devices offer a most promising new tech-
nique to increase the data-acquisition rate of planned satellite
experiments.
IMAGE SYSTEMS
Soft x-Rays
The soft X-radiation flux from different parts of the solar sur-
face varies considerably. Therefore, it is important to make
spatially resolved measurements of the X-ray flux so that a com-
plete intensity pattern over the whole surface can be obtained at
any instant of time. Problems of angular resolution, solar flux
levels, and detector area limit the system performance ; but con-
siderable effort is presently being put into this field, and it forms
an important part of the OS0 program.
The refractive index of soft X-rays in all familiar materials
is less than unity, and they are easily absorbed so that convcn-
tional mirror and lens imaging systems cannot be used. Re-
flection coefficients for soft X-rays are negligibly small except
20
 . INSTRUMENTATZON

I at grazing incidence angles of less than 1'. Imaging systems

of three kinds are possible-pinholes, zone plates, and reflec-
tion optics used at grazing incidence.
A pinhole camera with filters and photographic emulsion was
used to obtain the first soft X-ray photographs of the Sun (ref. 25)
and later photographs were taken by the Leicester group.
' Fresnel zone plates have been suggested for use in the soft X-
I ray region (refs. 26 and 27) but are extremely small and diffi-
cult to make and are nonachromatic. Their advantage over
the pinhole for soft X-rays is a shorter telescope length, but
potentially they are of greater use in the ultraviolet.
The technology of reflection optics for soft X-rays is still in
its infancy. Reflection is fairly high for a number of materials
for grazing incidence angles of less than 1" (ref. 28). Such
reflectors are usually tubular, the mirror profile being accu-
rately machined on the inside of the tube. Various surface
I
shapes have been considered, but single-reflection surfaces have
severe astigmatism and are only suitable as collimated light
I gatherers when used with a limiting aperture. Some double-
reflection profiles are stigmatic and form good X-ray images
l
over small collection angles. An X-ray telescope was proposed
by Giacconi et al. (ref. 29), and is shown in figure 2. This
~

field is being studied extensively by NASA (ref. 30). X-ray

~
reflection telescopes of both the collimating and image forming
types feature prominently in the O S 0 and AOSO programs.
Mechanical collimation systems can also be used with soft
X-rays, but they require scanning to build up a complete X-
ray picture and are, therefore, more suitable for satellite than
rocket vehicles. To obtain high sensitivity, multiple collima-
tion apertures are required, together with large-area detectors.
In principle, their mechanical simplicity enables them to com-
pete with the more complex reflection telescopes, but the latter
are already being developed greatly.
I
Ultraviolet
Since astrophysical studies in the vacuum ultraviolet can now
be carried on from outside the atmosphere, it is of interest to
consider the possibility of obtaining solar images in this spec-
I
I 21
 SOLAR PHYSICS

One-mirror system

/
A Grazing incidence

Axis of symmetry

Two-mirror system

Object Axis of symmetry

.
Image \
without coma

Figure 2 . 4 p t i c s for gratingincidence X-ray telescope.

tral region, suitable for television transmission. T o form the

first optical image at the camera tube in the vacuum ultraviolet,
the radiation should not traverse an absorbing medium. In
addition, it is desirable to image selectively within the camera
tube’s sensitive range. Thus a possibility for a suitable optical
system is a zone plate constructed without a substrate (ref. 3 1) .
Such a zone plate has been investigated by Baez (ref. 26)
for application to visible and ultraviolet image formation. It
consists of an array of concentric rings which are alternately
22
 INSTRUMWTATION

transparent and opaque. The radii of the ring boundaries in-

&ease as the square root of successive integers. At a distance
from the zone plate proportional to the square of the outside
radius of the largest transparent ring, and inversely propor-
tional to the illumination wavelength, light from every other
half-period zone is blocked off, so that all arriving light is in
phase and self-reinforced. This position is the primary focus.
The first plate consisted of a set of thin circular gold bands
made self-supporting by radial struts, leaving the transparent
zones empty. The diameter was 0.26 centimeter and it con-
tained 19 opaque zones, the narrowest of which was about 20 P
across. Tests were made in visible light and in the ultraviolet
only as far as 2537 A. Resolution and speed of the zone plate
are greatly improved over the pinhole. The plate compares
favorably in resolution, but not in speed, with a simple lens, if
comparison is made in the visible region. Below 1000 A, how-
ever, the speed is millions of times greater than that of a lens
made of any known material. The zone plate promises to be a
useful focusing device for the extreme ultraviolet since the reflec-
tivity and transmissivity of known materials makes lenses or mir-
rors extremely inefficient in this region.
SPECTROMETERS
soft x-Rays
In the soft X-ray region, wavelength resolution can be ob-
tained as follows :
( 1) By selective absorption in a filter composed of one or more
selected materials or by selective response in the
detector
( 2 ) By proportional counters with pulse-height analysis
(3 ) By Bragg reflection from a crystal, either flat or curved
While the grazing incidence spectrograph can resolve wave-
lengths well down into the soft X-ray region, it is currently used
primarily in the vacuum ultraviolet region and will be consid-
ered in that section.

209-916
23
 SOLAR PHYSICS

Selective Absorption

Selective absorption occurs in detectors and windows, either

intentionally or otherwise, and in many cases is used intentionally
to obtain wavelength discrimination. The first wavelength-re-
solved measurements of solar X-rays were made by NRL with
rockets. By using beryllium, aluminum, and Mylar windows
(ref. 9 ) , counters have been made sensitive to the wavelength
regions 1 A-8 A, 8 A-20 A and 44 A-60 A, respectively (see also
fig. 1) . The use of composite windows and comparison of the
signals from counters with different windows enable smaller
wavelength regions to be isolated, and the use of selected counter
gases such as argon (which has a K edge near 4 A ) can give a
selective response. These wavelength bands are broader, how-
ever, than is desirable and cannot be refined sufficiently. Also,
the rapid variation of sensitivity over the passband of the com-
posite filters and detectors makes interpretation of results
difficult.
During the early pioneer work in the solar soft X-ray field
and during subsequent work up to 1962, this type of detector
has provided most of the existing information on the solar
spectrum distribution. An indirect method of obtaining wave-
length resolution has been used by the NRL group (ref. 32)
and a Russian group (ref. 33) on rocket flights. This tech-
nique relies on the selective absorption of the atmosphere
over the rocket trajectory ; the harder X-rays penetrate to
the lower altitudes, while the softer components increase in
intensity with altitude. Hinteregger has employed the same
effect to analyze the composition and structure of the upper
atmosphere.
Proportional Counters
Using photoionization counters in the proportional region
considerably improves the wavelength resolution. High reso-
lution can be obtained at the short wavelength end of the
X-ray spectrum, but at wavelengths greater than 20 A too
few ion-pairs are produced to obtain satisfactory resolution.
Hence these counters have been used primarily to map the short
wavelength end of the spectrum and cannot be used to detect
24
 I

IN~RUMWTATION

’ 05 identify line spectra. Much development work on soft

X-ray proportional counters has been done by groups at
Leicester and London Universities. They have been flown
on Skylark rockets and on the Ariel I satellite (ref. 15) and
1 feature prominently in the O S 0 program.
Bragg Crystal Spectrometers
The crystal spectrometer is similar to that used for hard
X-rays (using a reflection crystal set at the “Bragg angle”),
but it uses crystals with relatively large lattice spacings of
the order of 10 A. They have high resolution, comparable
to that of grating spectrometers in the ultraviolet and will
resolve emission-line spectra. Such spectrometers (actually,
scanning monochromators) have already been used to observe
soft X-rays from laboratory plasmas (ref. 34) but have only
recently been developed for solar observations. The reflec-
tion coefficient of the crystals is usually small for soft X-rays,
and a high degree of collimation is necessary to obtain good
spatial and wavelength resolution. Certain organic crystals
such as potassium hydrogen phthalate enable measurements
to be made to 25 A and combine a reasonable reflection
coefficient with a uniform crystal structure. As this partic-
ular field is in its infancy and considerable importance is
attached to wavelength resolution, much effort is being cur-
rently applied to advance the technology.
A Bragg crystal spectrometer has been developed at NRL
and flown on an Aerobee rocket. This and a spectrometer de-
veloped by NASA (ref. 35) will be flown on OS0 satellites.
The NASA 1400-A spectrometer developed by Goddard
Space Flight Center actually consists of three separate Bragg
crystal monochromators and an ultraviolet grating spectrom-
eter. The crystal monochromators cover the ranges 1-3.8 A,
3.6-8.4 A, and 8.2-25 A. Parts of this instrument are shown
in figure 3. The detectors for the first two ranges are ioniza-
tion counters while the third is a Bendix open photomultiplier.
Future efforts in the crystal spectrometer field include exten-
sion of the technique to longer wavelengths, up to about 100 A,
25
 SOLAR PHYSICS
.

lncoming
radiation

Figure ~ . - ~ r y s t aspectrometer
~ for 1 A 4 . 5 A.

thus insuring complete overlap with grazing incidence spec-

trometers. For this purpose the development of crystals having
large lattice spacing is being studied.
Bent-crystal spectrographs, analogous to conventional Row-
land-circle spectrographs in the visible and ultraviolet, off er the
possibility of improved light collecting efficiency over the mon-
ochromator. They have been employed recently on rocket
flights using photographic film to record the dispersed spectra,
but they will require the development of X-ray image convert-
ers for use on unmanned satellites.
Recently (1964) the Los Alamos group (H. V. Argo, J. A.
Bergey, W. D. Evans, B. L. Henke, and M. D. Montgomery)
have developed and flown bent-crystal monochromators on
Nike-Tomahawk rockets. The equipment had a pointing ac-
curacy of about 295". A series of fixed-wavelength monochro-
mators were flown using crystals of mica and films of organic
26
 INSTRUMWTATION

materials including lead stearate. Resonance lines such as C V,

C' VI, N VI, N VII, 0 VII, and 0 VI11 in the wavelength
range 16-40 were isolated and detected by thin-window gas-
flow Geiger counters.
Ultraviolet
N o d Itrcidence Wauelengtb Regior

All spectrographs designed for rocket flight have used a re-

flection-type diffraction grating as the dispersing element. Be-
cause of the interest in the extreme ultraviolet spectrum, lenses
have generally not been employed, and focusing is by reflection
only.
Work with such instruments began in 1946, but it was not
until the very active period during the 1950's that most ad-
vances were made. A number of instruments were designed
and flown during these years; however, the stray light produced
by the intense long-wavelength portion of the solar spectrum
always set a detection limit on the weak emission lines in the
extreme ultraviolet. Improved measurements have been made
more recently and are described in a review by Tousey (ref.
36). The first good results were obtained with the 1959 flight
of Purcell, Packer, and Tousey (ref. 37) which succeeded in
showing a number of lines in the Lyman series of hydrogen and
the Lyman continuum. The photospheric continuum was also
detectable to approximately 1700 A, below which it was
swamped by instrumental stray light. These results were rec-
ognized as being about the best that could be obtained with a
single-dispersion system.
The greatest improvement in the normal-incidence spectro-
graph was achieved with the doubly dispersing instrument of
Detwiler et al. (ref. 38) shown in figure 4. The new spectro-
graph was essentially the normal incidence instrument used in
previous NRL experiments, with an added predisperser grat-
ing, external to the slit, instead of a collecting mirror. The
predisperser grating is of 40-centimeter radius of curvature,
having 600 lines/mm. The dispersion is arranged to be along
the slit; thus the visible and near ultraviolet portions of the
 SOLAR PHYSICS

15oooA ’ 5
Final

500 A
SI ~t and
dirk spectrum
Sunlight
y”

grating
(rhopad 10 neutralize
ortlgmatirm)

Figure 4.-The double-dispersionspectrograph (ref. 38).

spectrum fall beyond the slit and do not enter the spectrograph.
The dispersion directions of the two gratings are at right an-
gles so that the final spectrum is tilted, causing only a slight
difficulty in using a densitometer.
The double-dispersioninstrument was used first in April 1960,
and twice again in 1962. Resolution of 0.1 A was obtained in
one case. Because of the nearly complete elimination of stray
light, it was possible to follow the continuum down to nearly
800 A. In addition, many lines previously unresolved were
separated and some profiles measured.
Grazing Zncidence Wavelength Region
For many years it has been known that wavelengths near
100 8, or less can be resolved with diffraction gratings using near-
grazing incidence. The first solar spectrograph using a grazing-
incidence grating was flown in 1952 by Rense (ref. 3 9 ) . Several
improvements were then made and in a 1959 flight the instru-
ment was successful in detecting for the first time the resonance
28
 INSTRUMEh'TATION

Higb Resolving Power Spectrograpbs

Determining the true profile of a spectral line requires a resolv-
ing power of lo" to lo6. T o approach this with instruments de-
signed for flight, high orders of interference must be used.
Figure 6 shows a normal-incidence instrument (ref. 43) using
the 13th order, designed to obtain the profile of the hydrogen
Lyman-alpha line. Both gratings have a radius of curvature of
50 centimeters and are ruled with 1200 lines per mm, providing a
29
 SOLAR PHYSICS

Entrance slit-

Endless steel band

w i t h cutout s l i t s

Pigure 5.-The AFCRL grazinginci-

dence scanning monochromator.

Grotins
prsdi rperrer

&
/
Solor Image tn
Grating used
in 13th order
hydrogen Lyman-olpho

H i g h dispersion
spectrum o f hydrogen
Lyman-alpha

Sun1 ,gl.t

Figure 6.-NRL.'s stigmatic high-resolution Lyman-alpha line spectrograph.

30
 INSTRUMENTATION

dispersion of 0.4 A / m . Theoretically the instrument has a

resolving power of 250 000. In practice, however, the resolving
power is limited to 40 000 by the slit width, which was equiva-
I
lent to an instrumental width of 0.03 A. During two flights,
excellent line images were obtained, from which it was possible
to derive the emission line profile, as produced in the Sun and also
as modified by absorption in the Earth's atmosphere.
To obtain extreme resolution in the range of 3000 A to
2000 A, the spectrograph shown in figure 7 was built at NRL
using transmission optics. The echelle, with 73 lines/mm, pro-

Figure 7.-The echelle spectrograph for the 3000-A-2000-Arange.

duces a high dispersion at right angles to its rulings, in orders

between 120 and 60. The dispersion produced by this instru-
ment varied between 1.5 A/mm at 3000 A to 1 A/mm at
2000 A. Because of the stray light problem, a small mono-
chromator was used as a predisperser. Generally, the instru-
ment performed very well and produced spectra having
approximately the same resolution as is attained with a 21-foot
grating spectrograph. McAllister has developed an echelle
spectrograph based on this principle for rocket observations of
the Sun.
Another echelle instrument designed primarily for extreme
ultraviolet radiation is shown in figure 8. This was constructed

31
 SOLAR PHYSICS

Figure 8.-An extreme ultraviolet echelle spectrograph for hydrogen Lyman-alpha

line.

(ref. 44) using only reflecting optics and was intended for the
hydrogen Lyman-alpha line. Only a few narrow spectral ranges
could be covered, each approximately 4 A wide. In the labora-
tory the instrument was adjusted for a resolving power of
0.007 A. The first flight in 1961 failed because of pointing
difficulties.

32
 Chapter 4
Mupping the Sun's Ultraviolet Spectrnm

NORMAGINCIDENCE SPECTRAL REGION

I N TWO RECENT REVIEWS,Tousey (ref. 45) and Allen (ref. 46)

describe certain features of the solar spectrum and this in-
formation's contribution to an understanding of solar physics.
They also point out that the study of short-wavelength radiation
has just begun and that many, more precise observations are
required.
Within the 3000-A to 2085-A range, the Sun's spectrum is, as at
longer wavelengths, a continuum with Fraunhofer lines. More
than 6400 lines have been observed, although at least half of them
cannot be identified from the multiplet tables. At present, only
a small fraction of the data from this region has been published.
The lines of greatest interest here are the H and IS lines of Mg
11,at 2802.698 A and 2795.523 A respectively, which produce a
deep depression in the spectrum extending 20 A or more to either
side of their centers. The continuum level is also of great interest,
although earlier results showed no regions between 3000 A and
2085 A where the true continuum could be seen. With spectra
of 0.15-A resolution, several places between 2641 A and 2700 A
were reported where the continuum seems to be free from ab-
sorption and lies near the 5900" K level. Echelle spectra sup-
port this, but at shorter wavelengths there appear to be no win-
dows through which the continuum can be seen.
Between 2100 A and 2000 A, a sudden change takes place in
the spectrum. The intensity of the maxima between Fraun-
hofer lines falls sharply, and below 2085 A it follows a 5000" K
curve ;this is a decrease by a factor of nearly 4 in less than 25 A.
However, no abrupt change occurs in the bottom of the lines,
which continue to follow a 4900" K curve. T h e appearance is

33
 SOLAR PHYSICY

one of blanketing by continuous absorption. The continuum is

depressed, as though optical depth unity lay farther out in the
photosphere where the temperature is 5000" K. Furthermore,
the cores of the Fraunhofer lines seem to be formed at a higher
level where the temperature is lower.
The radiation in the 2000-A to 1200-A spectral range repre-
sents the transition region between the photosphere and the
chromosphere. The gradual transition from Fraunhofer spec-
trum to emission line spectrum takes place in this range so that no
Fraunhofer lines can be seen below 1530 A. Generally, the re-
gion between 500 A and 1500 A is represented by a continuum
and emission lines, both of which originate in the chromosphere.
The continuum is smooth except near hydrogen Lyman-alpha,
where it builds up toward the line center from both sides. The
most prominent emission lines in this region are those of the
Lyman series of hydrogen. The other lines are primarily those of
the more abundant light elements: He, C, N, 0,Ne, Al, Si, and S.
Most states of ionization, from neutral to 11 times ionized, are
represented. Agreement between photographic and photoelec-
tric data in this spectral region is excellent. Fluctuation in the
photoelectric measurements is regarded as noise and does not
generally interfere with line identification. A problem inherent
in the photoelectric technique is, of course, that scanning reduces
the amount of energy available in a given time so that it is diffi-
cult to obtain resolution as good as that of photographic measure-
ments for the same experimental conditions. Photography is
therefore expected to be of continuing value.
A brief review of the most significant flights has been given
recently by Hinteregger (ref. 47). The first photographic evi-
dence of some important lines in the extreme ultraviolet was
obtained from University of Colorado flights in 1958 and 1959.
Identification was difficult, however, because of a rather high
background of stray light and certain film defects. The published
list of observed wavelengths (ref. 40) attracted much attention,
but it also led to considerable controversy, which was subsequently
resolved by data from later flights.
The 1959 normal-incidence measurements made by NRL
(ref. 37) apparently realized the full potential of the single-
34
 MAPPING T H E S U N S ULTRAVIOLET SPECTRUM

I $ispersion photographic system. The photospheric continuum

appeared detectable down to about 1700 A, below which it be-
came swamped by instrumental stray light. Many emission lines
and a number of the hydrogen Lyman series were clearly recorded
' in these spectra, however. The data provided the first record
of coronal emission of Ne VI11 at 770 A and 780 %i, as well as
I
resonance lines of He I, N I, N 11, N 111, 0 11, 0 111, 0 IV,
0 V, Mg X, and S VI. The first indication of the hydrogen
Lyman continuum was also observed.
The most notable step in reducing the stray light problem was
taken with the introduction of the double-dispersion normal-
incidence spectrograph by Detwiler, Purcell, and Tousey (ref.
38) which was flown in 1960 and again in 1962. Spectra ob-
tained in 1960covered the range from 1550A to about 500 A with
a dispersion of 41 A per mm. Spectra with a dispersion of 20 A
per mm for the 2000-1200-A region and 10 A per mm for the
~

1250-800-A region were obtained with two similar instruments in

1962.
Most of the emission lines in this region are chromospheric.
For example, the first 13 lines of the Lyman series of hydrogen
were resolved in the 1960 spectrum, and were followed by the
~
Lyman continuum, with radiation temperature approximately
6650" K. He I and He I1 were noted, and the spectrum of C I
was present with great intensity and completeness. Further de-
tails of these spectra are given in the review paper by Tousey (ref.
45).
I
The AFCRL telemetering monochromators were flown several
times during the period 1959-61, producing a fair number of
records from 1300 A to 250 A (ref. 48). Various emission lines
were identified in this range and certain conclusions drawn from
the spectra, particularly that obtained on the August 1961 flight.
Among other things, the absence of lines from intermediate stages
of ionization indicated that the transition between the chromo-
sphere and corona is extremely sharp. It was also inferred that
the chromospheric lines (corresponding to ionization potentials
less than 100 eV) are emitted in a region of temperature less
than 60 000 OK, so that the higher series members are weak. It

35
 SOLAR PHYSICS

further appeared that ionization equilibrium in the corona is not

much different from what had been expected.
The best resolution of the solar spectra was obtained in the
range 3500 8, to 2200 8, with the echelle spectrograph of Purcell,
Garrett, and Tousey (ref. 49) flown in 1961. A resolution of up
to 0.030 A was obtained, from the 81st order at 3000 8, to the
122d order at 2000 8,. Orders were separated by crossing the
echelle with a CaFZ prism. The spectra from 3500 8, to 2988 8,
were of the same quality as spectra obtained from the ground by
conventional astronomical techniques. From 3330 A to 3350 A,
the echelle spectra agreed closely with the atlas compiled by
Bruckner at Gottingen. Within the 2988-8, to 2200-8, range, the
spectra contained some 4000 Fraunhofer lines. Wavelengths
were established to approximately 0.010 A by reference to the
many easily identified lines of Fe I, Fe 11, and other metals.
GRAZING-INCIDENCE SPECTRAL REGION

The first grazing-incidence data were obtained within the early

series of AFCRL telemetering monochromator flights during the
period 1959-61. The flights of January 1960 and August 1960
(ref. 42) , using a continuously moving exit slit and signals of the
position of the slit along the Rowland circle, yielded photoelectric
data down to 60 8,. The monochromators were operated with a
fast wavelength scan, resulting in a special resolution lower than
might be expected. These spectra are largely of historical
interest since they have been superseded by improved data. They
did, however, illustrate a conspicuous dropoff of spectral intensi-
ties below 170 A, leading to controversy as to whether this was a
real characteristic of the extreme ultraviolet spectrum or an
instrumental effect.
These data were compared shortly thereafter with NRL photo-
graphic records obtained with a grazing-incidence spectrograph
during June 1961 and August 1962 (ref.41 ). The wavelength
coverage of these flights was from 700 8, to approximately 170 8,
in first order. The lower cutoff was caused by an aluminum
filter which was used to suppress stray light and which became
extremely opaque at this point; however, had shorter wave-

36
 MAPPING THE SUNT ULTRAVIOLET SPECTRUM

lengths been present with great intensity they might have been
iecorded in second order. Since this was not the case, the earlier
discovered dropoff below 170 A was neither really confirmed nor
contradicted. Above 170 A the agreement between the photo-
electric and photographic records is entirely satisfactory. About
50 solar emission lines were observed with wavelength precision
of 0.1 A, by using He I1 303.78 A as standard. Themost
interesting of the identified lines were those of C VI at 182.35 A
and Fe XV at 284.3 A. Above 340 A the photographic spec-
trum is largely second, third and fourth order, and the only first-
order line which was definitely present was Mg IX, at 369 A.
The next important data were obtained when two Aerobee-Hi
rockets were launched in May 1963. The first (May 2 ) was a
scanning, grazing-incidence monochromator flown by Hintereg-
ger and colleagues of AFCRL (ref. 50) using a reduced scanning
rate and thus resulting in a greatly improved signal to noise ratio.
Replacing the tungsten photocathode with LiF deposited on a
glass substrate extended the long-wavelength-blind range and
increased the quantum efficiency at shorter wavelengths. The
spectra extended far beyond the previous limit of 170 A, and
lines were detected down to 65 A. The spectrum still showed a
sudden dropoff below 170 A; therefore the tungsten cathode
could be ruled out as a cause.
On May 10, 1963, the grazing-incidence spectrograph of Aus-
tin, Purcell, and Tousey (ref. 41 ) was reflown with decreased slit
width and a higher speed diffraction grating. Approximately
125 lines were clearly present above 170 A, and some 50 lines
(though not all resolved) between 45 A and 170 A. In this
region the data differ from the photoelectricrecord, probably due
to the different spectral efficiency characteristics of the instru-
ments.
The extreme ultraviolet emission lines which have been identi-
fied in the records from these flights have recently been tabulated
by Zirin (ref. 51). The following criteria were used for
identification:
( 1) Coincidence within 0.1 A
( 2 ) A strong and probable transition in an abundant ion

37
 SOLAR PHYSICS

( 3 ) Observation of other strong lines from the ion, in approx-

imately the right intensity ratios.
INTERPRETATION

Spectral Line Intensities

This section is concerned primarily with the emission-line solar
spectrum as obtained by rocket flights during the past 6 years.
Satellite data and additional interpretation will be considered
later.
An excellent review of these data has been provided by Pot-
tasch (ref. 5 2 ) . Emission lines from 1900 8, to about 15 A are
considered. The lines whose intensities are known are due
principally to ions with ionization potentials ranging from 13 eV
to 600 eV. These lines come from a region in the solar atmos-
phere where the temperature changes from about 15 000” K to
several million degrees. This is the “transition region,” which
has been only poorly understood from the past optical measure-
ments. For convenience the observed lines which have been
identified are reproduced in tables I and I1 (from Pottasch’s
review). A few belong to a stage of ionization not listed and
whose energy levels are at present unknown.
Table I.-0 bserved Emission-Line Zntensities
[From ref. 521

Intensity, ergs/cm*/sec
Probable
Wavelength, A identification
Oct. 10, Aug. 23, July 10,
1946 1961 1964

1892.03. . . . . . . . . . . . . 0. 08 .......... .......... Si 111

1862.78 . . . . . . . . . . . . . .04 .......... .......... A1 111
1854.72. . . . . . . . . . . . . .02 .......... ..........
1817.42 . . . . . . . . . . . . . .45 .......... .......... Si I1
1808.01.. . . . . . . . . . . . .15 .......... ...........
1670.81.. . . . . . . . . . . . .08 .......... ........... A1 I1
1657.00.. . . . . . . . . . . . . 16 .......... ........... CI
1640.47 . . . . . . . . . . . . . .07 ........... ........... He 11
1561.40.. . . . . . . . . . . . .09 ........... ........... CI
1550.77. . . . . . . . . . . . . .06 ........... ........... c IV
1548. 1 9 . . . . . . . . . . . . . .ll ........... ...........
1533.44.. . . . . . . . . . . . .041 ........... ........... Si I1
1526.70 . . . . . . . . . . . . . .038 ........... ...........

38
 MAPPING THE SUN’S ULTRAVIOLET SPECTRUM

’
Table I.-Obserued Em’ssios-Line I&em&ies-Gmthd

I Intensity, crpg/cm~/scc
Probable
~ Wavelength,A identification
Oct. 10, Aug. 23, July 10,
1946 1961 1964

1402.73. . . . . . . . . . . . . . 0.013 ............ Si IV

1393.73.. . . . .
1335.69. . . . . . . . . . . . . . ............. c 11
1334.52... . . . . . . . . .050 ........................
1306.02 . . . . . . . . . . . . . . .............
i 1304.86... . . . . . . . .
1302.17... . . . . . . . . . . .
1265.04... . . . . . . . . . . . ............
0 1

Si I1
1260.66... . . . . . . . . . . .
1259.53... . . . . . . . . . . . .002 ........................
1253.80.. . . . . . . . . . . . . s I1
1250.50.. . . . . . . . . . . . . s I1
1242.78. . . . . . . . . NV
1238.80... . . . . . . . . . .
~ 1215.67. . . . . . . . . H LY
1206.52.. . . . . . . . . . . . . .030 .OS(?) .071 Si I11
1201.0... . . . . . . . . . . . . .................
1194.2. . . . . . . . . . . . . . . .0015 ........................ s I11
1190.2. . . . . . . . . . . . . . . .0015 ........................
1175.70... . . . . . . . . . . . .010 .03 - 042 c I11
1085.71... . . . . . . . . . . . .006 .007 .009 N 11
1073.00. . . . . . . . . . . . . . .ooO8 ........................ s IV
1062.67... . . . . . . . . . . . .0008 ....................
1037.6. . . . . . . . . . . . . . . .025 .02a .025 0 IV
.020 .040 .036
1025.72.. . . . . . . . . . . . . .060 .050 .045
.OlO .01 . ......
1 989.79... . . . . . . . . . . . . .006 . . . . . . . . . -007
977.03. . . . . . . . . . . . . . . .050 .08 .081 c I11
972.5. . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ .01 1 H (3)LY
~ 949.74... . . . . . . . . . . . . .OlO .008 .005 H (4) LY
944.52... . . . . . . . . . . . . .0013 .0015 . . . . . . . . . . . . s VI
937.80. . . . . . . . . . . . . . . .005 .005 .004 H (5)LY
933.4. . . . . . . . . . . . . . . . .0029 .0024 ........ s VI
834. . . . . . . . . . . . . . . . . . . . . . . .016 .013 0 I1
0 111
790.10. . . . . . . . . . . . . . . . . . . . . . . . . . . .013 .009 0 IV
787.71 . . . . . . . . . . . . . . . . . . . . . . . . . . . .012 .008 0
786.48.. . . . . . . . . . . . . . .003 . . . . . . . . . . . . ...... sv
780.3. . . . . . . . . . . . . . . . .0065 .004 Ne VI11
770.4. . . . . . . . . . . . . . . . . . . . . . . . . . . . .013 .011
765.14. . . . . . . . . . . . . . . . . .0077 .006 N IV
760 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .004 ............ ov
718.5.. . . . . . . . . . . . . . . .003 . . . . . . . . . . . . 0 I1
....................... .m .007 0 I11

209-916 39
 SOLAR PHYSICS

Table 1.-Observed Emission-Line Intensities-Gmtinud

Intensity, ergs/cml/sec
_-___ Probable
Wavelength, A identification
Oct. 10, Aug. 23, July 10,
1946 1961 1964
~

686.34.. . . . . . . . . . . . .......... 0.003 .......... N I11

629. 73.. . . ... .......... .045 0.056 ov 1
625.3.. . . . . . . . . . . . . .......... .013 .......... Mg x
. 609.8.. . . . . . . . . . . . . ... ... .025
599.60.. . . . . . . . . . . . .......... .003 .......... 0 I11
584.33.. . . . . . . . . . . . .......... .055 .053 He I
.......... . 001 .. ... Ne V
569. . . . . . . . . . . . . . . . .......... .0025 .......... Ne V
... .0015 .......... Ne VI
558.60.. . . . . . . . . . . . .......... .0015 ..........
554. . . . . . . . . . . . . . . . .......... . 01 .......... 0 IV
550 . . . . . . . . . . . . . . . . .......... .0012 ....... Al XI
542.07.. . . . . . . . . . . . .......... .002 .......... Ne IV
537.02. . . . . . . . . . . . . .......... .0074 .......... He I
525.80. . . . . . . . . . . . . .......... .0017 .......... 0 I11
.......... .OlO .......... Si XI1
515.61.. . . . . . . . . . . . .......... .002 .......... He I
.......... .003 .......... 0 I11
.......... .021 .......... Si XI1
465.2. . . . . . . . . . . . . . .......... .013 .......... Ne VI1
......... .......... .007 ....... Mg VI1
430 . . . . . . . . . . . . . . . . .......... .0045 .......... Mg VI11
.......... .0075 .......... Fe XV
.......... .002 ....... Na VI11
40 1 . . . . . . . . . . . . . . . . .......... .008 .......... Mg VI
392.4. . . . . . . . . . . . . . .......... .OOl .......... A1 IX
385.0. . . . . . . . . . . . . . .......... .001 ..........
368.1 . . . . . . . . . . . . . . 0.032 .......... .031 Mg IX
365.9. . . . . . . . . . . . . . .012 .......... .......... Fe XIV
......... .018 .......... .......... Fc XVI
356.1 . . . . . . . . . . . . . . .009 .......... ....... Si X
.Ol .......... ..........
348.9. . . . . . . . . . . . . . .013 .......... ..........
.007 .......... .......... Si X
.007 .......... .......... Fe XIV
339.5. . . . . . . . . . . . . . .005 .......... ......
.028 .......... .......... Fe XVI
332.9. . . . . . . . . . . . . . .002 .......... .......... AI X
.003 .......... ..........
319.8. . . . . . . . . . . . . . .004 .......... ..........
315 . . . . . . . . . . . .Ol .......... ..........
308.6. . . . . . . . . . . . . . .......... .00045 ..........
.......... ..........
.004
303.78. . . . . . . . . . . . . .24 .25
.25 He I1
.......... .......... a. 025 Si XI
-The Si XI line was separated from the neighboring He I1 line b , he NRL in a flight
-
on May 10, 1963.
40
 MAPPING T H E SUN'S ULTRAVIOLET SPECTRUM

Table L-Observed Emission-Line Intensities-Cominued

Intensity, ergs/cma/scc
Probable
Wavelength, It identification
at.10, Aug. 23, July 10,
1946 1961 1964

........... 0.004 .......... SiIX

292.8 . . . . . . . . . . . . . . . .0025 .......... Si I X
2M.3. . . . . . . . . . . . . . . .021 .......... Fe XV
281.1 ............... .oO06 ..........
278.7. ........ ........... .oO06 .......... P XI1
277.05 .............. ........... .003 .......... Si VI11
272.0. .......... .003 .......... Si X
266.38. . . . . . . . . . . . . . ........... .OW7 .......... N V
264.4. .............. .004 .......... sx
259.3 . . . . . . . . . . . . . . . ........... .003 ..........
256.32 . . . . . . . . . . . . . . .015 .......... He I1
253.8 ............... ........... 002
1 .......... Si X
251.8. .............. ........... .0037 ..........
249.1 ........... ........... .0015 .......... Si VI
246.0. .............. ........... .0020 ..........
244.91. . . . . . . . . . . . . . ........... .002 .......... c IV
243.03 . . . . . . . . . . . . . . ........... .0037 .......... He I1
241.7. .............. ........ .0075 ..........
238.57. . . . . . . . . . ........... .0017 .......... 0 IV
237.33 . . . . . . . . . . . . . . ........... .002 .......... H e I1
232.5. . . . . . . . . . . . . . . ........... . m 3 .......... He I1
225.0. . . . . . . . . . . . . . . ........... . . . . . . . . . . 0.007
216.92. . . . . . . . . . . . . . ...... .......... .007 Si VI11
211.7 ............... ........... .......... .009
203.9 ............... ............ .......... .008
............ .......... .014
198.6. . . . . . . . . . . . . . . ............ .......... .OOlL s VI11
195.3. ........ ........... .......... .04
193.6 . . . . . . . . . . . . . . . ........... .......... .025
188.5. .......... ........... .......... -05
182.4. . . . . . . . . . . . . . . ........... .......... .017
180.6. .............. .......... -09
177.3. . . . . . . . . . . . . . . ........... .......... .08
174.7 . . . . . . . . . . . . . . . .......... .09
173.1 . . . . . . . . . . . . . . . ........... .......... .m 0 VI
172.17.. . . . . . . . . . . . . ........... . . . . . . . . . . .003 ov
171.3 . . . . . . . . . . . . . . . ........... .......... .09
154.0. . . . . . . . . . . . . . . ........... .......... . 004: ov
150.12 . . . . . . . . . . . . . . ........... .......... .m1 0 VI
148.3. . . . . . . . . . . . . . . ........... .......... .017
115.8. . . . . . . . . . . . . . ........... .......... .003: 0 VI
103.6. . . . . . . . . . . . . . . ........... .......... .m Fe IX
 .
SOLAR PHYSICS

Table 11.-List of Resonunce Lines

[From ref . 521

Ion Wavelength. b Isoelectronic f-value Intensity.

sequence ergs/cmS/sec
-~

c I1 . . . . . . . . 1335.7; 1334. 5 Boron . . . . . . . . 0.32; 0.32 0.10

c 111. . . . . . . 977.0 Beryllium. . . . . 0. 74 0.08
c IV. . . . . . . 1550. 7; 1548.2 0. 1; 0.02 0. 06; 0. 11
N I1 . . . . . . . . 1085 Carbon . . . . . . . 0. 70 0.009
N 111 . . . . . . . 991.5; 989.8 Boron . . . . . . . . 0.27; 0.27 0.012
765. 1 Beryllium. . . . . 0. 59 0.0077
N V........ 1242.8; 1238.8 Lithium . . . . . . . 0.08; 0. 17 0.002; 0.004
833 Nitrogen . . . . . . 0. 75 0.016
0 111. . . . . . . 834 Carbon . . . . . . . 0. 61 0.016
0 IV . . . . . . . 790. 1; 787. 7 0.22; 0.22 0.013; 0.012
o v........ 629. 7 0. 48 0.045
0 VI . . . . . . . 1037.6; 1031.9 Lithium . . . . . . . 0.07; 0. 135 0.014; 0.026
Ne IV . . . . . . 542 Nitrogen . . . . . . 0. 36 0.002
Ne V . . . . . . . 570 Carbon . . . . . . . 0. 6 0.0035
Ne VI . . . . . . 562. 8; 558.5 Boron . . . . . . . . 0. 17; 0. 17 0.003
Ne VI1 . . . . . 465. 1 Beryllium . . . . . 0. 47 0.013
Ne VI11 . . . . 780.3; 770.4 0.05; 0. 105 0.006; 0.013
Na VI11 . . . . 411. 1 0. 45 0.002
Mg VI . . . . . 400 0. 36 0.008
Mg VI1 . . . . . 432 Carbon . . . . . . . 0.45 0.01 1
Mg VI11 . . . . 430; 436 Boron . . . . . . . . 0. 28 0.011
368. 1 Beryllium. . . . . 0. 4 0.032
Mg X . . . . . . 625.3; 609.8 Lithium . . . . . . . 0.045; 0.085 0.013; 0.025
A1 111. . . . . . 1862.8; 1854. 7 Sodium . . . . . . . 0.61; 0. 32 0.04; 0.02
392. 4; 385. 0 Boron . . . . . . . . 0. 25 0.002
332.9 Beryllium 0. 35 0.002
550.0 Lithium . . . . . . . 0.08 0.0012
Si 11. . . . . . . . 1817.2; 1808.0 Aluminium . . . . 0. 6; 0.6 0.45; 0.15
Si I11. . . . . . . 1206.5 Magnesium . . . . 1. 62 0.049
Si IV . . . . . . . 1402. 7; 1393. 7 Sodium . . . . . . . 0.28; 0.57 0.013; 0.03
Si VI . . . . . . . 249.0; 246.0 Fluorine . . . . . . . 0. 34 0.003
Si VI11 . . . . . .319.8 Nitrogen . . . . . . 0. 15 0. 004
Si X........ 356. 1; 347.4 Boron . . . . . . . . . 0.09; 0.09 0.009; 0.007
Si XI . . . . . . . 303.4 Beryllium. . . 0 . 39 0.03
Si XI1 . . . . . . 521. 1; 499.3 Lithium . . . . . . . 0.035; 0.075 0.01; 0.021
278. 7 Beryllium. . . . . . 0. 38 0.006
s 11. . . . . . . . 125MO Phosphorus . . . . 1. 1 0. 004
1 190-1 202 Silicon . . . . . . . . 1. 13 0. 0045
s IV . . . . . . . . 1073.3; 1062. 7 Aluminium . . . . 0. 8 0.0016
s v. . . . . . . . . 786.5 Magnesium. . . . 1.4 0.003
s VI . . . . . . . . 944.5; 933.4 Sodium . . . . . . . 0. 29; 0. 58 ). 0014; 0.0027
s VI11 . . . . . . 198.6 F1uorine . . . . . . . 0. 13 0.0014
sx ... 264.3; 259.4 Nitrogen . . . . . . 0. 12; 0.08 0.004; 0.003
Fe 1X. . . . . . . 103.6 Argon . . . . . . . . . ............ 0.0094
Fe XIV . . . . . 366; 344 Aluminium . . . . 0. 1 ; 0. 15 0.012; 0.007
Fe XV . . . . . 284. 3 Magnesium . . . . 1. 1 0.032
Fe XVI . . . . 361. 7; 336.6 Sodium . . . . . . . 0. 16; 0.29 0.018; 0.028

42
 MAPPING THE SUN'S UL27UVZOLET SPECTRUM

' In general, the energy in these lines may be produced by col-

lisional excitation or radiative recombination. In the transition
region, however, collisional excitation is the important process
since radiative recombination is much slower, by a factor of up to
,
lo' for resonance lines. The rate of excitation by collision
becomes LNeNz, where L (=u",,) is the collision rate per elec-

I tron and ion, v the electron velocity, N e the electron density, and
Nz the density of the Z stage ion. Since the Sun's outer atmos-
phere is almost completely ionized, N e is related to the hydrogen
density NH by the approximation
Ne= 1.2"
The degree of ionization can be expressed by (Nz/NE) and the
abundance by (NE/"), where N Eis the total density for the
element E in all stages of ionization.
At electron densities below 101'/cm3, the principal means of
recombination from an ionized state is by radiative recapture
rather than by three-body collision. Thus the ratio of the popu-
lation of a single stage of ionization N,,, to that of the immedi-
ately lower stage N , is given by

where C,.,, is the rate of collisional ionization and (Y is the re-

combination coefficient. A useful set of ionization equilibrium
calculations has been made by House (ref. 5 3 ) , who treated all
elements from H to Fe and all ions from Z = 1 to 20.
In analyzing the above observations, it is assumed that each
collisional excitation from the ground state of an ion to the low-
est excited state from which an electric dipole radiation is allowed
results in the emission of a resonance-line quantum. The inten-
sity E (erg/cm*/sec ) of a resonance line emitted from the Sun is

where hv is the energy of a single quantum, C is the collisional

excitation rate, N is the ground state density of the ion in question,
R is the solar radius, and the integration is performed over the
43
 SOLAR PHYSICS

entire volume V above the solar limb. After rearrangement, the

quantity

emerges as an important indication of the atmospheric structure,

where the “height” h is an element of volume divided by the area
of the Sun. When this quantity is plotted as a function of tem-
perature for various elements, the curves are essentially the same
except that they are displaced from one another along the ordi-
nate. Pottasch (ref. 52) plots this expression versus temper-
ature for all the ions listed in table 11, after the relative abun-
dance of each element has been adjusted so that as little scatter
as possible from a common curve is obtained. The scatter is gen-
erally within 40 percent, with no differences of isoelectronic
sequences greater than this amount being observed.
The relative abundances obtained in this way, when compared
with the photospheric abundances found by Goldberg et al. (ref.
54), show no difference greater than a factor of two for half the
elements. For the four metals Mg, Al, Si, and Fe, this analysis
gives a greater abundance than does the photospheric analysis.
Line Profiles
Another set of data treating the same region and obtained with
rocket-borne equipment is that resulting from the study of emis-
sion line profiles. Profiles of the first two lines of the Lyman
series of hydrogen were observed in 1959, 1960, and 1962; the
1959 data pertained to a quiet region of the active Sun and that
of 1962 to a quiet-Sun condition. Self-reversal is deeper for the
quiet Sun, and clearly the lines are formed in a region where the
optical depth is large.
At present the only theoretical discussion of the hydrogen
Lyman-alpha profile is that due to Morton and Widing (ref. 55) ,
which is based on a study of the transfer of radiation in a simple
two-level atom. Naturally many approximations had to be
made, and it was necessary to assume that the line was formed in
an atmosphere of constant density and with only a guess as to the
temperature gradient. Further, no velocity fields were used.
Under these conditions, a profile was made to fit the observed
44
 I MAPPING T H E SUN’S ULTRAVIOLET SPECTRUM

shape in the core of the line successfully, but the fit was very poor
& the wings. The fit is obtained by adjusting the temperature
and density where the line is formed so that the separation of
peaks and absolute intensity can be matched. For the normal
regions of the Sun the electron density and temperature found are
l N .=3 X 1Os/cm3 and Te=90 000 K. This does not seem com-
O

1 pletely reasonable since to obtain an optical depth of 10’ for the

center of hydrogen Lyman-alpha at this density, a scale height
of 500 000 kilometers is needed, an unreasonably large number.
Obviously, the interpretation of the Lyman profiles requires fur-
ther study, and radio-frequency observations may provide some
of the information required for interpretation.
Structure of the Quiet Solar Atmosphere
The physical problems of the Sun’s outer atmosphere have
been separated into almost independent studies of the chromo-
sphere and the corona. The pressure and temperature differ-
ences between these layers are so large that any attempt to inter-
polate between them usually fails. The situation is made worse
~
by ambiguities concerning both the top of the chromosphere and
the bottom of the corona. Zirin and Dietz (ref. 56) have sug-
gested that the individual spicules provide most of the optical
~

radiation from the top of the chromosphere and that the regular
chromosphere is not observed higher than about 2000 kilometers
~
from the base of the chromosphere. On the other hand, there
are no observations giving the height of the lowest part of the
corona. All that is known is that it is below 10 000 kilometers.
It can be said with some certainty that the transition layer
must be quite thin. This conclusion is based on the absence of
all the intermediate ions of Ne, Mg, and Si, with ionization poten-
tials between 100 and 300 eV. Since only maximum values for
the intensity of these “missing” lines can be given, we can infer
that the transition region, having a temperature between 50 000”
l and lo6 OK, has less than one-tenth as many atoms as the
corona or the higher temperature portion of the chromosphere.
I However, the fact that spicules and prominences are features of
this region makes further analysis difficult. At present many dif-
I 45
 SOLAR PHYSICF

ferent models are being tried in an effort to explain the observa-

tions.
Centers of Activity and Flares
The variations in extreme ultraviolet radiation and its corre-
lation with solar activity make it clear that a considerable portion
of this emission comes from active areas of the Sun. It is also I

clear that these changes can be divided into two types: (1 ) slow
variations with a life of several days, probably thus influenced by
solar rotation, and ( 2 ) sudden variations associated with flares.
It is the slowly varying emissions that come from solar active cen-
ters, which may be plages or coronal condensations. Although
the geometry of active areas is likely to be complex, little else can
be done but to regard them as horizontally stratified, with out-
wardly increasing T and decreasing Ne. Data required for
calculations and analyses of these regions is rapidly improving
and much more will be done in the near future.
Flares introduce additional problems. In the first place,
they are more variable in shap, time, and character than other
features. Secondly, the detailed line spectrum of flare emission
below 10 A has not yet been observed. In addition, the electrons
in flares may not have a Maxwellian velocity-distribution. Until
emission lines characteristic of flares ( < 10 A) are observed, the
best hope of progress is to calculhte emission from assumed models
and compare it with available unresolved spectroscopic observa-
tions. Such a flare analysis has been made by Pounds et al. (ref.
5 7 ) . By interpreting the spectral distribution as thermal emis-
sion, they obtain flare temperatures up to 10' OK. However, by
assuming the radiations to be dominated by line emissions at the
long wavelength end of the observed region, higher temperatures
could be calculated. Certainly much further development is
needed for treatment for all of these features.
Laboratory Astrophysics
One of the primary difficulties in the analysis of the solar ex-
treme ultraviolet spectrum is the proper identification of spectral
lines. Laboratory wavelengths, intensities, and identifications
are not available for the vast majority of the lines observed.
Theoretical identification often requires uncertain extrapolations
46
 MAPPING THE SUN’S ULTRAVIOLET SPECTRUM

from the last well-calculated or observed energy level separation.

Improvements in the techniques of spectroscopy, image fonna-
tion, and photometry for this spectral range would be of great
benefit.
Values of oscillator strengths and collision cross sections are
also required for all atoms and ions whose spectral lines are pres-
ent. It is not sufficient to know these parameters for just the
immediate levels from which the spectral lines arise. Accurate
rate processes must be known for all transitions important in
establishing the equilibrium configurationsof the atom. Labora-
tory measurements and theoretical calculations must be extended
to supply these data.

. 47
 *
chapter 5

Mapping the Szcn's S o p X-Ray Spectram

T HE COMPLETE MAPPING of the solar X-ray spectrum is a

complex problem, for it is necessary to obtain simultane-
ously wavelength and spatial and temporal resolution of radiation
whose flux is low. During the past 6 years measuring techniques
have gradually improved : the early detection and crude wave-
length resolution measurements using the newer techniques have
determined the shape of the short wavelength end of the spectrum
and demonstrated the hardening of the spectrum in the presence
of flares and solar disturbance. Pinhole camera measurements
have shown that relatively small areas of the solar surface are
responsible for most of the X-ray flux,and other measurements
have shown the variable nature of the flux. Satellite measure-
ments have enabled the flux variations to be observed over long
periods and correlated with simultaneous measurements in the
visible and ultraviolet regions, with radio bursts, and with iono-
spheric phenomena.
PREDICTION AND DISCOVERY

Historically, interest in extraterrestrial X-radiation arose from

attempts to understand the mechanism of maintenance of the
E-layer in the ionosphere, since energies of at least 12 eV are
needed to ionize the atmospheric constituents, energies which
cannot be explained on the basis of a Planckian radiator at the
photospheric temperature.
Hulbert (ref. 58) and Vegard (ref. 59) suggested that solar
X-rays might be responsible for the E-layer ionization. Edlen
(ref. 60) identified certain coronal lines as being caused by
elements in highly ionized states and Hoyle and Bates (ref. 61)
estimated that a flux of between 10" and lo-' erg/cm'/sec would
be required at wavelengths of about 40 A to account for the
E-layer.
49
 SOLAR PHYSICS

The first positive identification of extraterrestrial soft X-rqys

was provided by Burnright (ref. 4) using photographic emd-
sion and filters flown on captured V-2 rockets. Further flights
with different filters showed that X-rays were present in the
8-16-A region and beyond 111 A, limitations being set by absorp-
tion edges in the filters.
Friedman and his colleagues (ref. 62) at NRL flew photon
counters and showed that an appreciable X-ray flux exists at
wavelengths less than 10 A but is negligible below 7 A. Further
rocket measurements in the 1952-58 period were made under
various solar conditions by the NRL group (ref. 9 ) using photon
counters and ionization chambers with filter windows. The fluxes
found were as follows :

( 1) <8 A, to 1.5x lo-' erg/cm'/sec

( 2 ) 8 A-20 A, about 10-serg/cmz/sec
( 3 ) 40 A-100 A, about lo-' erg/cm'/sec
Too few measurements were made to obtain reliable informa-
tion about the variation with solar conditions, although more
variation is apparent below 8 A than for the longer wavelengths.
For a review of this work, see also Friedman (ref. 63).
THEORIES OF THE X-RAY FLUX

Elwert Theory
The solar soft X-ray flux at middle wavelengths can be de-
scribed by a blackbody distribution at temperatures around lo6
O K , but it has absolute values many orders of magnitude less

than for a Planck radiator. Elwert (ref. 64) made theoretical

estimates of the solar soft X-radiation by considering the free-
free and free-bound bremsstrahlung and excitation-recombina-
tion radiation from coronal electrons. This gives continuum and
line spectra, and Elwert considered the radiation from the various
constituents of the corona including highly ionized elements, at
the short wavelength end. By comparing the relative intensities
of visible and ultraviolet coronal spectral lines, estimates were
obtained of the coronal temperature from the known line transi-

50
 .
MAPPING T H E SUN‘S SOFT X-RAYSPECTRUM

’
’ tipn probabilities. Temperatures of the order of 10” OK were
obtained. The ionization equilibrium, electron densities, and
radiation spectrum were calculated. Order-of-magnitudeagree-
ment with experimental flux levels was obtained. Acton has
criticized Elwert’s calculations on the basis of revised abundan-
I
ces, inadequacy of the Born approximation for collisional excita-
I tion cross sections, and the contribution of recombination to the
ionization equilibria.
1 synchtoaon Radiation
Stein and Ney (ref. 65) have calculated the continuum flux
resulting from the synchrotron process in an attempt to estimate
the radiation of flares at all wavelengths. They found that rea-
sonable magnetic fields and electronic energy spectra similar to
the observed proton spectra could account for the radiofrequency
and visible continuum measurements. Under these conditions,
however, synchrotron radiation is negligible at soft X-ray wave-
lengths. Very high fluxes of very energetic electrons would be
required to produce any significant synchrotron X-ray emission.
Inverse Compton Effect
I
The scattering of low energy photons by relativistic electrons
has been suggested by Schklovsky (ref. 66) to account for the
hard component of solar flare X-rays. I n an interaction known
as the “inverse Compton effect,” the electron loses energy to the
photon, changing the latter’s effective wavelength. Schklovsky
calculated the electron energy distribution necessary to account
for the observed X-ray fluxes and found it in agreement, within
an order of magnitude, with that estimated from the optical spec-
tra of flares. Acton has criticized this theory on the basis of
balloon observations of the hard X-ray flare spectrum.
Fluorescence Radiation
The classic case of fluorescence radiation involves the ioniza-
tion of an atom by removal of an inner shell electron by bombard-
ment with energetic electrons. A fluorescence transition occurs
when an outer shell electron drops into the inner shell. Acton
(ref. 67) calculated fluorescence radiation for Fe, S, and Ne, and
51
 SOLAR PHYSICS

found it negligible in comparison with thermal radiation ; hoy-

ever, the fluorescence mechanism may become significant in some
nonthermal processes as, for example, in a plasma excited by 15
keV electrons.
Bremsstrahlung
Bremsstrahlung is the radiation emitted when energetic elec-
trons are decelerated by collision with dense matter. The
apparent drift velocities of electrons exciting type 111 radio
bursts correspond closely to the energies of hard X-rays observed
rarely with certain flares. Of course the radio bursts originate
in upward-directed streams, while bremsstrahlung must result
from electron impact at chromospheric and photospheric levels.
Microscopic processes for solar X-ray emission have not been
investigated in any detail. There is considerable uncertainty
in the values of opacity and cross sections used in calculations of
the above mechanisms and in the ionization equilibrium calcu-
lations upon which these are based. Nevertheless, the basic proc-
esses are sufficiently well understood to serve as guides in the
initial interpretation of solar X-rays observations.
FORMATION OF THE IONOSPHERE

The soft X-ray spectrum of the quiet Sun covers the region
from about 10 A to 100 A and is absorbed in the Earth’s atmos-
phere in a fairly well-defined region between 100- and 150-
kilometer altitude, forming the E-layer of ionization. The
shorter wavelength radiation penetrates farthest, but the ex-
ponentially increasing atmospheric density produces a fairly
well-defined lower altitude limit to the ionizaton profile. During
disturbed conditions higher energy X-rays are produced which
lower the level of the ionized region. That the soft X-ray flux
is primarily responsible for the ionized E-layer has been demon-
strated by observation of the change in critical frequency for radio
wave reflection from the ionosphere during disturbed conditions.
The observed change, by a factor of 1.5, is in good agreement
with a value of 1.6 calculated for the known maximum and mini-
mum values of the solar X-ray flux. Furthermore, analysis of
the ion composition of the E-layer indicates that the equilibrium
52
 MAPPING THE S U N S SOPT X-RAY SPECTRUM

ions are those produced by X-rays rather than by ultraviolet

;adiation.
ECLIPSES AND SUDDEN IONOSPHERIC DISTURBANCES

, Observation of the critical radio frequency during an eclipse

indicates an inhomogeneous X-ray source distribution assuming
I that the E-layer is due to X-rays. The ionospheric eclipse is
irregular because of masking and unmasking of the discrete
sources of ionizing radiation. Rocket measurements during the
eclipse of October 12, 1958 (ref. 25 ) ,showed 10 to 13 percent of
X-ray flux remaining at totality. The X-ray intensity of the
corona was six times greater in the west limb, which contained
several active plages, than in the east limb. Although it has been
shown that the D-layer is produced mainly by the hydrogen
Lyman-alpha radiation (ref. 68), this fact does not adequately
~
explain the height decrease of the base of the D-layer during
sudden ionospheric disturbances. Earlier rocket measurements
showed enhancement of X-rays only during flares. Measure-
ments by the NRL-SRI satellite have confirmed that the soft
X-rays and not the hydrogen Lyman-alpha cause the flare SID
events.
~ The most effective wavelengths in producing the D-layer are
those less than 8 A, and these may be increased by many orders of
~

magnitude during flare events while total X-rays only double.

Flare spectra have been measured (ref. 63), and the hardening
of the radiation is probably due to suprathermal electrons (ref.
64) since the gas is cool. If the radiation were thermal, a tem-
perature of lo8 O would be required to explain the observed
radiation and the -alpha line would not be emitted. The short
wavelength limit during flare enhancement probably lies in the
region of the Fe K series between 1 A and 2 A. It was also
demonstrated that the ionized layer at 70- to 75-kilometer alti-
tude, below the normal D-layer, would result from line emission
at about 2 A rather than from continuum.
ROCKET MEASUREMENTS
During the early observations by the NRL group before 1958,
crude measurements were made of the X-ray spectrum in three
53
 SOLAR PHYSICS
-
bands by using different counter gas fillings and windows (ref.9) .
Variation of the X-ray flux with solar conditions was also noted.
During the following 6 years, knowledge of the soft X-ray flux
has been increased by a number of rocket flights, notably by the
NRL group, the Leicester and London University groups, and
a Russian group.
These measurements have mapped the X-ray spectrum in more
detail, in particular at the short wavelength end, and flights un-
der various solar conditions have enabled a start to be made on
the difficult, but important, process of correlating X-ray flux
with visible solar phenomena. Cutoff of the flux below about
10 A is observed under quiet solar conditions, but the flux in this
region increases by many orders of magnitude during flares, al-
though the increase is much less at a wavelength of 50 A. (See
fig. 9.) During flares the short wavelength limit extends down
to about 2 A and a strong correlation exists between the solar
activity and the flux at short wavelengths. At sunspot minimum
the solar X-ray flux has a total value of about 0.1 erg/cm'/sec
and can be fitted to the Planck distribution for a temperature of
5 X lo5 OK. The flux increases to about 1 erg/cm2/sec at sun-
spot maximum under quiet conditions.
In April 1960 the NRL group obtained the first X-ray pinhole
camera picture of the Sun (ref. 5 ) from an Aerobee rocket, giving
a resolution of about 0.1 solar diameter with a pointing accuracy

I I I 1

Figure 9.-Solar X-ray emission for

various solar conditions.

54
 ~~ ~~ ~~

MAPPING T H E SUN'S SOFT X-RAY SPECTRUM

of. 1 arc-minute. Accidental rotation of the camera during the

exposure through 160" about the pointing axis smeared the point
images into arcs. In spite of this smearing, the X-ray sources can
be seen to correspond with the bright plages on simultaneous
calcium K-line photographs. It was concluded that the X-ray
sources were smaller than the plage areas. The brightest X-ray
source had 70 times the background intensity, and less than 5
percent of the area of the hemisphere was responsible for 75 per-
cent of the X-ray emission. The mean solar diameter in X-rays
is about 6 percent greater than the optical diameter, a difference
of about 40 000 km.
A radioheliogram of the solar emission at 3300 Mc/sec, ob-
tained at Stanford University, is similar to the X-ray heliogram
when the former is rotated through 160" and photographed.
However, the X-ray photograph shows greater contrast, which
is interpreted as being due to enhanced X-ray line emission from
active coronal regions. The photograph also showed that most
of the X-ray emission originated in relatively small volumes of
condensed coronal material coinciding with magnetically active
plage regions at a higher temperature than the surrounding
atmosphere. Line spectral measurements between 10 A and
25 A were made by the NRL group (ref. 69) using a Bragg crystal
spectrometer carried in an Aerobee rocket. The spectrum was
uncollimated so that line widths gave a direct measure of the
angular width of the line-emitting region. Eighty-five percent
of the integrated flux in this region was due to emission lines.
Emission lines of highly ionized states were identified; namely,
Fe XVII, 0 VII, 0 VI11 and N VII. The narrowness of the
Fe XVII and 0 VI11 lines indicated that they originate in active
regions of the same dimensions as plages; the remaining lines are
broad as if from the whole solar disk.
The use of crystal spectrometers, together with the use of
grazing-incidence telescopes for obtaining solar images, is a
relatively new application in solar physics. Both techniques are
important for they extend the wavelength and spatial resolution
respectively by at least an order of magnitude over previous
methods. Crystal spectrometers have been developed by the
groups at Goddard Space Flight Center, at NRL, and at Leicester
55
2-18
 SOLAR PHYSICS
.
and London Universities, and have been flown on rockets ( 1964)
in their own right or as developmental projects for the O S 0 an'd
AOSO programs. Grazing-incidence optics have also been
rocket tested by the Goddard Space Flight Center (ASE) and
the Leicester and London groups. Further rocket flights are
scheduled for these instruments in 1965.

56
 I

chapter 6
Evidence of Sokzr H a r d X-Ray Emission

T HE FIRST OBSERVATION of hard solar X-rays was made by

balloon by Peterson and Winckler (ref. 70) in March 1958.
Solar flare X-rays of energy greater than 20 keV penetrate the
atmosphere to low altitudes. The flux was shown to be a burst
of less than 18 seconds’ duration and of energy in the region of
0.5 MeV. Subsequent measurements have been made by bal-
loon by Anderson, Winckler, Vette, and Casal of short bursts in
the 20- to 150-keV energy range. Ionization, Geiger counters
and scintillation counters have been used as detectors. The
NRL group (ref. 71) has shown that the spectral distribution
during a flare of importance 2+ fits a source temperature of lo8
O K and an electron density of the order of 1Og/cm3.

I n the March 1958 flare, an intense burst of 800 Mc/sec radio

emission occurred, simultaneous with the X-ray burst, but pre-
ceding the flash phase of the flare. I n a later event (August
11, 1960), the spectral distribution was close to that of the NRL
measurement, and coincided with type I11 radio-noise bursts.
The flare of September 28, 1961, was observed by Anderson and
Winckler (ref. 72) in the 20- to 150-keV range and showed a
“precursor” followed by a rapid rise and decay of flux,the decay
being broken into three stages of about 1-minute duration each.
Radio bursts at 9500 Mc/sec coincided with the X-ray pulse peak
and the precursor, while type I11 radio noise bursts coincided
with the breaks in the decay. The visible flare started 14 min-
utes before the X-ray flash and reached a maximum 8 minutes
after it. Microwave emission at 2000 Mc/sec followed the
H-alpha event quite closely, but had no distinct similarities with
the X-rays.
The extreme variability in the isolated balloon measurements
of flare characteristics showed that considerably more sampling
was necessary. Therefore, satellites have been used for sub-
57
 SOLAR PHYSIU

sequent measurements of the short wavelength end of the X-ray

spectrum.
There has been considerable speculation concerning the source
of the hard radiation. One suggestion is that it may be produced
by high energy protons exciting plasma oscillations, but these
protons would have to be of 100-MeV energy. De Jager has
proposed that the type I11 radio bursts are excited by streams
of electrons accelerated in the flare region and moving into the
corona with energies of 10 to 100 keV; he has proposed that the
hard X-rays are produced simultaneously. Kawabata found
that the microwave emission coincident with large flares can be
considered to consist of an intense single burst and a long-term
enhancement of flux called a “post burst increase.” He has com-
pared flare X-ray fluxes with thermal models and attributes the
2-8 A emission to thermal emission at a temperature of 2 X 10‘
OK, although for shorter wavelengths higher temperatures must
be postulated to obtain the observed spectral distributions. It is
necessary to combine both the thermal and electron bremsstrah-
lung component theories to more fully explain the observed flare
X-rays.

58
.
chapter 7
Corona

E NERGY IS GENERATED in the interior of the Sun, transported

to the surface, and then escapes. Were it not for a flux
of mechanical energy in the outer layers of the solar atmosphere,
the density would decrease rapidly above the photosphere and
the temperature would be no greater than the photospheric tem-
perature, about 5000" K. However, the mechanical energy
gives rise to shock waves and is eventually degraded to thermal
energy, causing the outermost layers to become very hot. This
is the origin of the corona, which may be observed out to several
solar radii during eclipses. The radiation from the corona is
only about one part in 10"of the total radiation from the Sun,
meaning that the corona is difficult to observe; however, observa-
tions of the inner corona may be made with a coronagraph. The
corona may also be observed at X-ray, ultraviolet, and radio
wavelengths. The interpretation of these observations is re-
viewed by Seaton (ref. 7 3 ) .
CORONAL. TEMPERATURE

The temperature of the corona, as deduced from ionization

equilibrium, is about 6 x 1 0 5 OK. At that temperature Fe X
is near maximum abundance and the higher states of ionization
are proportionately less abundant. Relative abundances have
sometimes been better explained by assuming the corona to be
thermally inhomogenous, the temperatures of quiet regions being
as low as 6 x lo5OK and disturbed regions as high as 1.2 X 10"OK
(ref. 63).
X-ray emission data have yielded temperatures as low as
5 X lo" at solar minimum. Measurements of the profile of the
Fe X red and Fe XIV green lines, however, generally give tem-
peratures greater than 2 x 10" OK if thermal broadening is as-

59
 .
I

SOLAR PHYSIC3

sumed. Some of the discrepancy may be attributed to the fact

that lower temperature measurements refer to the quiet mid-
mum corona and higher temperatures to more excited regions.
De Jager and Kuperus (ref. 74) , in discussing the mechanical
energy flux of the Sun, found that dissipation of shock wave
energy in a completely quiet Sun could not produce a corona tem-
perature higher than 6X lo5 OK. They concluded that the Sun
is never completely quiet and that local heating from magnetic
fields persists even at solar minimum, sufficient to increase the
average temperature by at least lo5 OK. To reach a coronal
temperature of 2 X 10" OK, however, would require a mechanical
flux at the coronal base equal to 2 percent of the solar constant.
Thus, microturbulence or AlfvCn waves may play a role in pro-
ducing the observed line profiles.

CORONAL GREEN LINE AND X-RAYS

The red (6374 A) and green (5303 A ) lines are emitted by the
ions Fe , X ( E ,=233 eV) and Fe XIV (E,=355 eV) , respec-
tively, and originate in the corona. The red line is fairly uni-
formly distributed over the corona, but the green line is concen-
trated in centers of activity associated with sunspots. Because
the green-line intensity has often been correlated with ionospheric
variability, it is of interest to compare its behavior with the ob-
served changes in X-ray emission. A comparison made during
1959 rocket measurements showed that a 13-fold increase in
green-line intensity corresponded to a 19-fold increase in 8-20-
A X-ray flux.
I t has been pointed out that over the interval from minimum
to maximum of the past cycle, the average green-line emission
varied by a factor of 8 to 10, due perhaps to an increase in both
temperature and density of the active regions. This green-line
variation correlates well with the sevenfold variation in total
X-ray flux. The red line changed very little but clearly showed
a maximum intensity at sunspot minimum. Presumably the
temperature of the corona decreased at solar minimum to a value
closer to the maximum excitation temperature of the red line.

60
c
CORONA

OS0 B CORONAGRAPH

The brightness of the sky prevents the corona from being seen
far from the limb with a coronagraph. If such an instrument
were placed above the Earth‘s atmosphere, it would, in principle,
make possible the observation of the white light corona at all
times. At balloon altitudes of about 30 kilometers, the sky is
reduced to a brightness nearly as low as that during a total solar
eclipse. Above 100 kilometers, except for possible airglow emis-
sions, the sky brightness is effectively zero, offering no impedi-
ment to rocket or satellite-borne instrumentation. Accordingly,
a Lyot coronagraph with an external occulting disk was designed
for the OS0 B satellite and tested on two Aerobee-Hi rocket
flights in June 1963. Two instruments (both designed at NRL
under contract to NASA) were flown at this time, one using
photoelectric recording and the other photographic. The field
covered by these instruments was from R=3.5 to 10 R-.
These were the first records of the solar corona brightness for
these values of R made without the aid of a total solar eclipse.
The results were far from perfect, but they are consistent and
show that considerable improvement will result when the use
of a larger vehicle will permit the external occulter to be placed
farther away. All of the photographs show greater brightness
in the direction of the Sun’s equator and the approximate direc-
tion of the Earth’s ecliptic than at right angles. Of course this
light is in part the zodiacal or F-corona. It appears probable,
however, that a portion is true K-corona since the brightness over
the Sun’s south pole was greater than over the north. Pictures
taken at the High Altitude Observatory during the eclipse of June
20, 1963, also show greater brightness over the south than the
north polar region. The zodiacal light, on the other hand, is
expected to be symmetrical from north to south.
The corona photographs were of no value closer to the Sun’s
limb than R =3.5 R, This distance is a little beyond the outer-
most part of the corona photographed during the July 20, 1963,
eclipse and resulted from the closeness of the occulter.
The photoelectric coronagraph completed two scans during
the time available, the first in tangential and the second in radial
polarization. The results are neither as smooth as one might
61
 SOLAR PHYSICS
-
expect, nor do they show any clear-cut features such as streamers.
The irregularities were partly produced by the electrical nohe
and partly by the foreign matter known from the photographs
to be flying through the field of view. However, the general
shape and the radial distribution are in agreement with the
photographic records.
However, the polarization data obtained photelectrically are
not in agreement with other work. At the inner edge of the
region covered, the tangential component was brighter than the
electrons. From R=5 to 11 R o ,however, the polarization was
in the opposite sense to that measured by others, although the
magnitude was not measured. Possibly this observation is not
correct, although such a condition has been suspected. At
present no explanation exists for this result.
All the data seem to suggest that the Moon can never quite be
matched as an occulter. Probably, however, with a multiple
disk occulter at 10 feet or more from the objective, an orbiting
coronagraph will be able to monitor the white-light corona with
at least as great resolution and sensitivity as can be realized from
the ground during a total eclipse.

62
 chapter 8

Satellite Monitoring of Solar Radiation

the X-ray flare. Moreover, the flight time of a rocket is not

63
 SOLAR PHYSZCS

Solar Radiation I
The first successful solar monitoring satellite, called Sol&-
Radiation I (SR I ) , was put into orbit in June 1960. To pro-
tect the X-ray photometers from charged particles, the NRL
group used permanent magnets to deflect the particles from the
detector window. SR I carried an ultraviolet photometer, an
X-ray photometer and a photocell solar aspect sensor. The
X-ray detector was an argon-filled ionization chamber with a
beryllium window which gave a passband between 2 8, and 8 A.
The satellite spin exposed the detector for short intervals while the
sensor monitored the satellite’s attitude relative to the Sun. Data
were recovered only by real-time telemetry to the NASA Mini-
track network and to a few isolated stations. Consequently SR I
provided solar observations only 1.2 percent of the possible time.
These were transmitted during more than 500 data-acquisition
passes between June and November 1960. Approximately 20
percent of these records indicated detectable X-ray flux in the
2 A to 8 A band. Apparently the shielding magnet was success-
ful in protecting the X-ray photometer from Van Allen belt
radiation, since no Van Allen belt modulation was detected in
any of the SR I data. However, the magnet strongly influenced
the satellite spin dynamics. The component of the permanent
magnetic field along the spin axis coupled with the geomagnetic
field to produce a torque which resulted in an 8-day precession.
The changing spin rate coupled with this precession resulted in
complicated satellite kinematics. Precession also affected the
attitude so that the Sun was under observation by the satellite
only 50 percent of the time that the satellite was in sunlight.
SR I provided the following significant X-ray measurements :
( 1 ) The quiet Sun normally does not emit more than 6X lo-’
ergs/cm’/sec below 8 A
( 2 ) Some solar activity was usually visible when a large X-ray
flux was observed below 8 8, in excess of the quiet-Sun
threshoId
( 3 ) Some ionospheric effect (SID) could be detected when
the X-ray flux in this band exceeded 2 X 10-3ergs/cm*
sec

64
 S A T E U I T E MONITORING OF SOLAR RADIATION

(4) Significant changes in the X-ray flux can occur in times

as short as 1 minute
(5) Other types of solar activity besides flares produce SID ;
included are active prominence regions, bright surges
at the limb, and subflares at the limb.
Solar Radiation III
I n 1961 the composite satellites SR I11 and Injun I were
launched. SR I11 had two ionization X-ray detectors for the
2-8-81 and 8-14-A bands, while Injun I (intended for energetic
particle studies ) had a mica-window Geiger counter originally
intended to measure low-energy geoelectrons in the radiation
belts. The counter also responded to the 2-14 X-ray band
with a sensitivity which could be determined approximately.
The satellites did not separate but spun slowly together in
orbit. However, it was possible to obtain data (Acton, ref. 67)
from the detectors and some attitude information.
The combined results of the SR I11 and Injun I satellites can
be considered in two separate categories: X-ray emission of the
quiet Sun and X-ray emission of solar flares. For the sake of this
discussion, Acton defined the quiet Sun as the condition prevail-
ing between impulsive outbursts of optical or radio flare activity;
in other words the X-ray intensity will, by definition, show no
large changes in level from one observation to the next, or from
one day to the next. The NRL pinhole photographs show that
individual active regions play an important role in enhancing the
total X-ray flux from the Sun. Thus, continuity of observations
is essential to validate any measurement on the quiet Sun so de-
fined. For this reason, nearly 60 percent of the observations by
this combined satellite were made during only a 5-day period.
Transient events are well distinguished from the slowly varying
background flux ; uncertainties relate to the differentiation be-
tween active region radiation and quiet-region radiation. One
important result of Acton’s analysis is that the microwave (2800
Mc/sec) radiation and the X-ray radiation correlate in gen-
eral, but not in detail. This is interpreted as the difference be-
tween the scales of source regions for X-rays and microwaves.
The latter are conjectured to be optically thin on a microscopic
65
 SOLAR PHYSICS

dale over an entire active center, though possibly optically thick

in smaller structures (less than 2 arc-minutes) . The microwate
radiation source, on the other hand, is thought to be comparable
in size to an entire active center because of the general distri-
bution of electron densities of critical magnitude. The micro-
wave emission is also less dependent upon small variations in
the local temperature and density than is the X-ray flux.
The total quiet-Sun X-ray flux on a quiet day is of the order
of 5 X erg/cmz/sec, ( X < 14 A) and the coronal temperature
must be about 2 . 3 ~ 1 0OK. ~ (The NRL pinhole photographs
of the Sun indicate that most of this X-ray emission comes from
the immediate neighborhood of active regions : coronal tem-
peratures can be inferred from coronal emission-line profiles.)
I t is assumed that all the X-ray flux is produced from optically
thick plasma at this temperature, isolated in the active centers
and defined by the extended calcium K-line plages. In other
words, Acton concluded that essentially all quiet Sun X-radiation
comes from coronal enhancements and streamers which overlie
the active regions roughly as delineated by the calcium K-line
plages. Using the estimated electron densities and the observed
extent of the emission regions, the intensity of 2800 Mc/sec radia-
tion can be estimated from such a coronal condensation. The
calculated flux is approximately 27 flux units, which is within
a factor of 4 of the observed magnitude.
The observed short-term variations of the X-ray flux do not
correlate neatly with variations in the microwave emission. One
mechanism suggested by Acton to account for this discrepancy in-
vokes high density regions of roughly cm3, at 2.3 X lo6 OK,
with electron densities of the order of 101’/cm3. Regions as
small as this could change their X-ray flux within the time scale
observed, but variations in their radiofrequency emission will
not be observable with existing equipment. Obviously the at-
tendant visible light phenomena will lie at the threshhold of res-
olution, and conceivably they have not been observed. Only
the “moustaches” appear dimensionally similar to such events.
The X-ray emission from coronal prominences has not been ob-
served by any of the spacecraft previously discussed. It is ob-
vious from the very high temperatures prevalent in these plasmas
66
 SATELLITE MONITORING OF SOLAR RADIATION

{asindicated by the Ca XV emission) that they must have intense

X-ray emission.
SR I11 data relate to a number of flare-lie events. For this
discussion it is convenient to distinguish between thermally emit-
ting flares, in which the X-ray emission results from local en-

i hancement of electron density and temperature, and “nonthermal

flares,” in which the electron velocities depart s i m c a n t l y from
a Maxwellian distribution and may be anisotropic. SR I11 ob-
I servations confirmed the previously well-established fact that
“thermally emitting” X-ray flares are more common. Such
flares produce minor ionospheric effects or none at all. Another
subclass of flares, those with an explosive phase, frequently pro-
duced nonthermal X-ray bursts. Studies on ionospheric effects
may validate or qualify these crude satellite data on the relatively
I
harder X-ray spectrum of the nonthermal flares. As a basis for
discussion, Acton constructed two thermal models and one non-
thermal model solar flare to provide comparison with the ob-
servational results. Acton concludes that both thermal and
nonthermal processes are important and necessary to account
for the optical, radio, and X-ray emission of flares.
Based on the radiofrequency and X-ray emission and depend-
ing also upon the relative roles of thermal and nonthermal proc-
esses and the energies of electrons involved, the following four
categories of flux distribution can be defined :
( 1) The coolest class of flares produce little if any X-ray or
radio emission.
( 2 ) Flares which produce small increases in X-ray intensity
without marked hardening generally emit no strong microwave
bursts and produce no notable ionospheric effects.
(3) Nonthermal radiation increases at wavelengths shorter
than 2 A, and thus the ionosphere disturbances become most
significant. When the nonthermal component is noticeable but
not at maximum intensity, microwave bursts will not be strong
and will usually rise and fall gradually in intensity.
(4) The most extreme nonthermal events produce hard X-
rays (greater than 20 keV) and X-ray enhancements up to
two orders of magnitude above the quiet-Sun level. Such strong

67
 SOLAR PHYSICS

bursts always produce ionospheric disturbances of the greatest

magnitude and are accompanied by microwave bursts of the
impulsive type.

Recently the SR VI, a further NRL solar radiation satellite,

has been put into orbit and is working satisfactorily.
os0 I

NASA's Orbiting Solar Observatory program was designed

to conduct experiments in solar physics, astronomy, and geo-
physics above the Earth's atmosphere from a platform pointing
at the Sun. The OS0 program consists of eight planned
launches. OS0 I was successfully launched on May 7, 1962,
and will be followed by OS0 I1 in 1965. A description of the
scientific results of OS0 I is given by Lindsay (ref. 75).
The OS0 orbit is circular at an altitude of 300 nautical miles.
The spacecraft (fig. 10) weighs 500 pounds and consists of a
lower wheel section of nine wedge-shaped compartments, four
containing control and recording equipment and five containing
30 pounds each of experimental equipment. This section rotates
at 30 rpm to provide gyro stability and is mounted below a semi-
circular sail section which is kept pointing at the Sun to within 2

Figure 1O.-Orbiting Solar Observatory.

68
 SATELLITE MONlTORlNG OF SOLAR RADlATlON

arc-minutes. The sail can support two sets of experimental

eiuipment weighing 50 pounds each. A more detailed descrip-
tion is given by Dolder et al. (ref. 3 ) .
For X-ray measurements, the pointing section of O S 0 I con-
tains a xenon-filled ionization chamber detector with a beryllium
window 0.005 inch thick, developed by GSFC and sensitive to
the range 1 A to 8 A (fig. 11 ) . Scintillometers developed by
GSFC, covering 0.5 1 MeV and 20-1 00 keV are also in the point-
ing section. In the wheel section are several gamma-ray detec-
tors covering the 50-keV-3-MeV and 100-500-MeV regions.
Results on slow variations and transient events in the 1-A-8-A
region have been reported by White (refs. 76 and 77), and the
observations of X-ray bursts in the 20-100-keV region by Frost
(ref. 78). A slowly varying component of 1-A-8-A flux was
1.0

”
%

.E._ 0.1
-
w

- Theoreticnl
X Experimentol

0.01
01 03 1 .o 3 10
Wavelength, A

Figure I 1 .--OS0 I ion chamber efficiency as a function of wavelength.

 SOLAR PHYSICS

observed which correlates well with the slowly varying compp-

nent of 2800 Mc/sec solar radiation (fig. 12). This radiation
can be accounted for by localized sources having the same hori-
zontal extent as calcium K-line plages with thicknesses propor-
1
tional to their d ameter and having an electron temperature of
about 2.8X lo6 OK and an electron density of about 5X log elec-
trons/cm3. A further conclusion is that the ratio of line emission
to continuum emission is at least 10: 1, and more probably 30: 1,
for these conditions.
In addition to a slowly varying component, transient events
(X-ray flares) lasting from 10 minutes to a few hours were fre-
quently observed (fig. 13). Correlations with H-alpha flares,
with sudden ionospheric disturbances, and with 2800 Mc/sec
transients have been investigated; the results show that as an in-
dicator of local solar activity, the OS0 I X-ray experiment was
more sensitive by a large factor than indicators based on iono-
spheric effects or than indicators based on observations of solar
flux in visible or radio wavelengths. X-ray flares were fre-
quently observed associated in groups possessing a characteristic
pattern.
Observations of the transient bursts of soft X-rays indicated
that they could not be directly associated with microwave ob-

IO-A solor
background

1,
1 6
I ,
12
,
18 24
(

M
,

5
,1962

11
1
17
1 - - L L L - -
23 29 5
I
II
I
17
J
23
March April MOY

Figure I2.-Slowly varying components of 2800-MC/sec solar flux and of 1-A-8-A

X-rays.

70
 SATELLITE MONITORlNG OF SOLAR RADIATION

servations because of the extreme disparity in sensitivity of the

tGo methods. However, the observations by Frost show a direct
correlation between the 20-100-keV X-ray bursts and the micro-
wave impulses together with sudden ionospheric disturbances.
Frost concludes that the high-energy X-ray bursts are produced
in the chromosphere by a nonthermal source.
Ariel I
The first joint U.S.-U.K. satellite Ariel I (or UK-I) was
launched in April 1962. It carried a proportional counter de-
tector with pulse height analysis giving a spectral resolution of
approximately 2 A in the 4-14-A range (ref. 14). I n the first 3
weeks of the l i e of this satellite, more than 150 measurements
were made of X-ray flux over the entire wavelength range, rep-
resenting averages over periods of 4 to 8 minutes, to a precision
of about 10 percent (ref. 15). Even when flares were definitely
known not to be occurring, on a few occasions the quiet-Sun
X-ray flux varied by as much as 75 percent during an hour or
less. There was some correlation between these small variations
in the total flux and terrestrial observations of various solar in-

oa
00

0 0

I I I I I I
0145 0150 0155 02W om5 0210
Uni*er.al ?,me

Figwe 13.-Development of soft X-ray burst recorded by OS0 I, March 8,1962.

71

209-916 0 - u
 SOLAR PHYSICS

dices. Under quiet-Sun conditions, the intensity in this band

decreased sharply at shorter wavelengths so that usually tke
two highest energy channels gave no reading at all. These
spectra correspond roughly to 1.5 x lo6 OK. The integrated
energies were estimated to be 3x10-3 ergs/cm'/sec for 8 A to
14 A, and 1.3X ergs/cm'/sec for wavelengths less than 8 A.
The values are more than 100 times greater than Elwert's esti-
mated thermal radiation for a million-degree corona. Since
Elwert's model was for an isothermal corona, the temptation is
to postulate hotter localized regions at about 2 million degrees.
The spectra, however, do not confirm this suggestion. Bowen
has made an alternative suggestion that the spectrum of the quiet
Sun in the 8-A to 14-A range is hybrid, consisting of a thermal
component upon which are superimposed the nonthermal or
quasi-thermal spectra of localized condensations.
Approximately 20 solar flares were observed in the X-ray
region during the first 3 weeks. Generally X-ray flare spectra
were confirmed not only as an intensification, but also as a hard-
ening of' the quiet-Sun spectrum. The magnitude of the flare
X-ray emission varied considerably and did not correlate well
with the importance of H-alpha flares. The high time resolution
of the Ariel I spectrometer (0.1 to 1 minute) revealed that flare
X-rays are emitted in brief flashes and do not occur as gradual
brightenings to a single maximum and subsequent fading as is
observed in the visible hydrogen spectrum. Flares exhibit con-
siderable variety in their spectra. Thus, one example revealed
gradients corresponding to a gray body at effective temperatures
ranging between 2 and 3.8X106 OK. Another well-observed
example showed the relationship between spectral change in the
flare X-rays and ionospheric responses. During the rise to maxi-
mum, the X-ray spectral distribution was essentially preflare in
character so that the change was of intensity rather than of dis-
tribution. However, at the peak of the flash phase of this flare
during the maximum of flare X-ray emission, the spectral dis-
tribution had hardened considerably. The spectral distribution
corresponded to a plasma in quasi-equilibrium whose tempera-
ture rose from 10 to about 12X 1O6 OK during a 3-minute period.
The X-ray flux at this phase was 2 x 10-3ergs/cm'/sec. Electron
72
I SATELLJTE MONITORING OF SOLAR RADIATION

densities necessary to account for this flux can be calculated

aGd correspond to those prevailing in the low corona or the
chromosphere.
Other Satellite Measuremenu
Soft X-ray measurements have been made from a Russian
satellite (ref. 17) using open-type photomultipliers as detectors
and inserting various filters in sequence to select wavelength
ranges. In this way various portions of the spectrum were
selected from 1 A into the ultraviolet. Measurements under
quiet conditions were consistent with a blackbody spectral dis-
tribution at 0 . 9 x lo6 OK, with little radiation below 1 A, while
under disturbed conditions the curve could be represented by
an additional distribution at a temperature of loe OK.
The Orbiting Geophysical Observatory, OGO I, launched in
~

September 1964 contains scintillation counters to measure high

energy solar X-rays.
Two X-ray monitoring satellities were launched in October
1963 as part of the Vela program. These are in high orbits,
permitting almost continuous monitoring from 0.5 A to 10 A.
Ten detectors are distributed over the surface of the spacecraft.
Pairs of filters give separate spectral sensitivity bands of 1-20
keV (0.5-10 A) and 3-20 keV ( 0 . 5 4 A ) . During intense flares
the flux measured by the thick filter detector was considerably
less than that passing the thin filter. Many X-ray events were
recorded, correlating in general with reported solar flares. How-
ever, on some occasions the observed flares exhibited no X-ray
emission, and at other times there were X-ray events with no con-
current reported flares.
C. S. Warwick, in a private communication, has compared
these Vela data with ground observations of X-ray effects upon
the lower ionosphere. Flare X-rays create a sudden increase in
electron density at heights corresponding to the D-layer and be-
low. This response increases ionospheric absorption of cosmic
noise (usually measured around 18 Mc/sec and called SCNA) .
Simultaneously, reflection of a VLF signal (say 16 kc/sec) be-
comes more efficient,and a sudden enhancement of atmospherics
(SEA) is observed.
73
 SOLAR PHYSICS

Warwick’s comparison showed that all ionospheric events

(SEA or SCNA) were accompanied by a measurable increase
in solar X-rays. However, only about one-third to one-fourth
of the Vela events were associated with reported ionospheric re-
sponse. Both the 0.5-10-A and the 0.54-A detectors correlated
with ionospheric responses when the flux exceeded well-defined
thresholds. Since NRL’s 1960 data showed a similar associa-
tion for 2-8-A X-ray flux while the 1961 data did not, Warwick
concluded that radiation shorter than 10 A is more important for
producing ionospheric effects than that in the 10-20-A range.
Warwick’s analysis answers the question : to what extent can
ground observations of X-ray events by their indirect effects sup-
plant satellite-borne direct observations? SCNA data alone pro-
vide as many false alarms as real indications, unless corroborative
data such as 3000-Mc/sec radio flux measurements are con-
sidered. About 80 percent of the bursts at these frequencies are
accompanied by major X-ray events, and SCNA qualification
would reduce the false alarm rate. For lesser X-ray events, no
ground observations of indirect effects are sufficient to replace
satellite measurements.

ULTRAVIOLET RADIATION

The N R L SR I Satellite
The first solar radiation monitoring experiments to be carried
out from an orbiting vehicle were those of the NRL SR I Satellite
(known also as Solar Radiation I, Greb I, and 196072). This
was placed in orbit on June 22, 1960, and produced useful data
for about 5 months.
The ultraviolet detectors were nitrous oxide ionization cham-
bers with LiF windows, which responded to radiation from
1000 A to 1350 A. Difficulties encountered with these hydrogen
Lyman-alpha detectors necessitated reliance on measurements
made from rockets during this period for absolute calibration,
and no hydrogen Lyman-alpha data were recorded after October
22, 1960. Records show that during the period from July 13
to August 3, hydrogen Lyman-alpha changes did not exceed 18
percent. The important result was the establishment of an
74
 SATELLITE MONITORING OF SOLAR RADIATION

upper limit on possible variation of hydrogen resonance line

intensity. During a disk flare of importance 2, the measure-
ments clearly showed an upper limit in the flare emission of 11
percent of the total disk quiet-time flux. Intensity of flare
hydrogen Lyman-alpha thus appears to be too low to play a
significant role in the D-layer phenomena.
os0 I
More extensive solar monitoring experimentswere made p s i -
ble by the launching of O S 0 I in March 1962. The spectrom-
eter used was a grazing-incidence grating instrument similar to
Hinteregger’s described earlier, having for the detector a Bendix
magnetically focused strip-type photomultiplier. This instru-
ment made continuous solar observations, whenever the satellite
was in sunlight, for nearly three solar rotations in March, April,
and May, 1962. During this period extreme ultraviolet spectra
were obtained at the rate of one every 8 minutes, or approxi-
mately 7 per orbit, resulting in an estimated 7000 spectra being
recorded for the first 2% months of operation. These observa-
tions were followed by nearly 1 year of intermittent observations.
The spectrometer was designed to cover the wavelength range
from 10 A to 400 A. However, below 100 A the decreasing
sensitivity, combined with an increase in scattered light made
distinguishinga reliable spectrum impossible. The most reliable
observations were thus obtained in the 150400-A region (ref.
79)’ and agree well with rocket observationsin this region. The
valuable additional information obtained demonstrates the use
of satellite observations in identifying spectral lines and in study-
ing changes in the solar corona with time, especially those changes
associated with active centers and flares.
Line Zdentifitation
One of the difficulties of line identification in the far ultraviolet
has been the richness of the emission-line spectrum. Making
observations over a period of time and searching for groups of
lines whose intensities have the same time dependence might
solve this problem. This suggestion was based on the expecta-
tion that lines of one multiplet arising from closely spaced upper
75
 SOLAR PHYSICS

levels will vary by the same amount with changes in solar activity,
thus imposing an additional constraint which line identifications
must satisfy.
This method was used in searching for the permitted transitions
of Fe XIV. Lines of this ion are expected to be grouped into
four multiplets between 200 A and 400 A. Examination of the
time variations of lines in this region shows three groups with
separations compatible with predicted splittings. Two of these
groups may well correspond to the 2p-2p0 and 2s-2p0 transitions
since they lie within 2 A of the computed wavelengths. Two
other lines, at 204 A and 21 1 A, have the same variation and
probably correspond to the 2d-2p0 transitions. No other lines
exhibit the same time dependence.
Having examined one set of lines, it is of interest to examine
other lines of the same element, which perhaps belong to dif-
ferent stages of ionization. For example, those lines common
to both the zeta discharge and the Sun may be iron lines. Es-
sentially four groups of lines have been observed by the OS0 I
spectrometer according to the amplitude of the fluctuations.
The group with somewhat less variation than the above Fe XIV
lines may well originate in Fe XIII, as suggested by the similarity
with those of another possible Fe XI11 line at 364 A. The re-
maining lines have been placed in two not completely distinct
categories. The least variation is shown by a group of lines near
170 A, and also by a possible Fe X line at 345 A, indicating that
these last two groups may originate in Fe X, Fe XI, and Fe XII,
the smallest variation being associated with FeX.
To summarize the observations, those ions existing at lower
electron temperatures generally vary less than those at higher
temperatures. Since the variation depends upon solar activity,
this will be mentioned before further discussions of these
observations.
Variations of the Extreme Ultraviolet Spectrum W i t h Solar Activity
The use of OS0 I has permitted the acquisition of a solar
extreme ultraviolet spectrum which can tentatively be associated
with a corona disturbed by varying centers of activity. The closest
approach to observation of a quiet Sun was obtained on March 11,
76
 SATELLITE MONITORING OF SOLAR RADIATION

1962, when the observed face of the Sun had been free of dis-
turbances for the preceding 6 months. An analysis of emission
lines, made for the period from March 7 to April 5, shows that the
lowest counting rates of the period were observed when the sun-
spot number was near zero and the calcium K-line plage area
was also at a minimum.
It was also clear, however, that no exact correlation could be
assumed to exist between the extreme ultraviolet flux and ground-
based observations. Fe XV and XVI lines, for example, were
lower on May 1 when the sunspot number was 49 than on May 15
when it was 15. Similarly, although the agreement between the
radio data obtained at 2800 Mc/sec by the NRC and extreme
ultraviolet coronal fluxes is sometimes remarkable, it is not truly
consistent. A small maximum at 2800 Mc/sec on May 1 to May
5 does not appear at 284 A, whereas on succeeding days a larger
peak is observed at 284 A than at radiofrequencies. The region
under observation in May when a small maximum was observed
at 2800 Mc/sec was a flare-producing center of activity roughly
1 solar rotation old, whereas the plages existing on the Sun on
May 9 to May 15 are for the most part remnants of active centers
four to five rotations old which were no longer sites of flare
activity.
These observations suggest the necessity of having knowledge
of the recent history of solar activity as well as current data to
correlate extreme ultraviolet radiation with other information.
The extreme ultraviolet emission lines each display individual
features which vary from one line to another. In particular
He line fluctuations, but not others, can sometimes be associated
with the brightening of existing plages and the occurrence of
radio noise storms at 169 Mc/sec as recorded in France.
The coronal lines of Fe XIV, Fe XV, and Fe XVI are strongly
associated with plages, but they appear to have residual intensi-
ties even if the Sun shows no sign of activity. Although large
fluctuationsoccur in the relationship of Fe XV to the plage area,
the Fe XV emission is more strongly associated with plages than
Fe XIV. Assuming that the regions of increased Fe XV emis-
sion are equivalent in area to the plages, a plage to quiet-Sun Fe

77
 SOLAR PHYSICS

XV ratio of between 200 and 300 to 1 may be calculated. This

quiet-Sun component may be associated with fine structure or
it may be uniformly distributed; better spatial resolution is re-
quired to determine this. If both radiations are assumed to
originate in the same region of the corona, a temperature of the
order of 1,75X 1O5 OK is obtained from the theory of ionization
equilibrium. This temperature is to be compared with a value
of 8X105 OK usually obtained for undisturbed portions of the
corona.
To summarize the data on other extreme ultraviolet lines,
those ions existing at electron temperatures below about lo6 OK
(lower stages of Fe, Si VI11 through Si X, and Mg VI11 and IX)
show little association with active regions, while those ions exist-
ing above lo6 OK show strong association with plages and active
regions. It is not clear whether certain large increases in Fe
XV and Fe XVI result from a combined increase of electron
temperature and density over plages or whether localized re-
gions in which these emissions might occur merely increase in
number over plages. Better spatial resolution is required to
determine how the smaller increases in intensity are distributed
over the solar disk.
Flare Activity
The OS0 I spectrometer observed transient events superim-
posed on the slow variations in the spectrum. Generally it is diffi-
cult to obtain reliable data for such events because the increase
in radiation is small compared with the radiation from the entire
disk and because the long time required to scan the spectrum (8
minutes) severely limits the time resolution for a partiaular
spectral line. However, several tentative conclusions can be
drawn from the data. In the first place, of all the extreme ultra-
violet lines examined, the largest increase during the flares is
a.ssociated with the ion having the highest ionization potential,
Fe XVI. Secondly, even for this spectral line the increase is
considerably less than for harder X-rays. For a flare of im-
portance 2+, on March 13, 1962, the Fe XVI line at 335 8, in-
creased by 60 percent, while the 1-10 8, ionization chamber out-
put (ref. 80) increased by a factor of more than 3.

78
 SATELLITE MONITORING OF SOLAR RADIATION

In general, the Fe XVI flux increases more than does that from
F6 XV, thus altering the 355 A/284 A ratio. In many cases no
change whatever is observed for the Fe XV flux during the flare.
The 304-A line shows an increase of about 10 percent approxi-
mately coincident with the optical flare. A 30 percent increase
l
is observed in the 335-A line, but only after a class I1 radio
burst has occurred. No change is observed in the 284-A line.
OS0 I also monitored the total flux of hydrogen Lyman-alpha,
I providing data for a period during which five flares of optical
' importance 2 occurred. All hydrogen Lyman-alpha enhance-
ments occurred mostly within the duration of the H-alpha flare.
Apparently, there is no typical time dependence of the varia-
tion in hydrogen Lyman-alpha. The time of peak hydrogen
Lyman-alpha enhancement does not necessarily coincide with
the H-alpha flare maximum, but usually precedes it slightly.
The flare of March 22,1962, provided the best observed hydro-
gen Lyman-alpha enhancement (ref. 81 ). It is unique among
all the events in that the time course of the hydrogen Lyman-
alpha enhancement was similar to that of a classic H-alpha flare.
The time of hydrogen Lyman-alpha maximum coincided exactly
with the reported optical flare maximum.

79
 chapter 9
The Future Flight Program

SOLAR EXPLORER SATELLITE

T HEIQSY SOLAR EXPLORER S A T E L ~ Eto be launched by

NASA late in 1965 will perform continuous monitoring
of solar X-rays. The orbit altitude will vary in the 650- to
1020-kilometer range, and the satellite will carry nine X-ray
photometers. In concept the satellite is similar to the NRL
Solar Radiation series, but it has facilities for data storage as
well as direct transmission. Also, there is provision for reorien-
tation of the spin axis so that the instrument viewing time is
kept to a maximum. Data transmitted continuously is intended
to be received by any convenient station.
ORBITING SOLAR OBSERVATORY

A total of eight launches has been approved for the O S 0 pro-

gram; the remaining schedule showing two in 1965, one in 1966,
two in 1967 and one each in 1968 and 1969. Payloads for
O S 0 B through O S 0 E have been selected as of January 1965.
O S 0 B and later spacecraft also include a Sun-scan raster 40
arc-minutes side and 1 arc-minute resolution. In the soft X-ray
range the selected experiments include, specifically, a crystal
spectrometer developed by NRL for the observation of solar
flares and grazing-incidence spectroheliographs developed at
GSFC/ASE and at Leicester and London Universities. In addi-
tion the flights will carry scintillometers developed at GSFC,
Geiger counters developed at NRL and Leicester and London
Universities, proportional counters developed at Leicester and
London Universities, and ionization chambers developed at the
University of Michigan and N U .
The crystal spectrometer and spectroheliographs will be used
to obtain detailed wavelength and spatial resolution of solar
X-radiation and thus increase the knowledge of the nature of the
81
 SOLAR PHYSICS

individual sources. The various ionization and scintillation de-

tectors will provide continuous monitors of the radiation in spe-
cific wavelength bands so that studies and comparisons can be
made over as large a variety of solar conditions as possible. Some
additional soft X-ray measurements are also being made as part
of the OS0 program.
ADVANCED ORBITING SOLAR OBSERVATORY

The AOSO (fig. 14) is intended to make continuous, detailed,

high-resolution studies of solar phenomena with considerably
higher pointing accuracy than the OS0 spacecraft. This space-
craft should enable much higher spatial resolution to be obtained
in measurements of solar phenomena such as flares. It has a
normal pointing accuracy of t 5 arc-seconds of the radiometric
center of the Sun, an offset point of similar accuracy within a
40 arc-minute square centered on the Sun, and two raster scans,
one coarse and one fine. A total weight of 250 pounds is avail-
able for experiments.
Four launches have been planned so far, one in 1969 and three
more during the next 3 years. Single launches following at
yearly intervals have also been proposed. The program also
involves the development of instrumentation for solar viewing in
various wavelength ranges, including a high-resolution ( 5 arc-
seconds) X-ray telescope by Lindsay and Giacconi (GSFC/
ASE) . At present, four experiments have been selected for the
project, and these are briefly considered below.
High-Resolution X-Ray Telescope
The object of the experiment with a high-resolution X-ray
telescope is to obtain X-ray spectroheliograms, particularly of
plage regions, with a resolution of a few arc-seconds. The in-
strument will cover wavelengths from 6 A to 60 A: the 6-lo-A
region should include radiation from flares, and the longer wave-
lengths are of interest for quiet-Sun and plage studies. The main
instrument will use double-reflection grazing-incidence optics
(fig. 2 ) with a large collecting area of several ten’s of square
centimeters. The X-ray image detector will consist of a thin
scintillator, an image intensifier, fiber optics, and a television
82
 THE FUTURE FLIGHT PROGRAM

b
To Sun

instrument
window

I
36 in. J

Figure 14.--Advanced Orbiting Solar Observatory.

camera. A square picture of 2.5 x lo5resolution bits is proposed.

A second, low-resolution (1 arc-minute) telescope will also be
used for fast time resolution of solar disturbances.
White Light Coronagraph
The White Light Coronagraph experiment (Eddy and New-
kirk, HAO) is designed to record changes in brightness, extent,
and polarization of the outer corona, both long term (days to
months) and short term (minutes), with angular resolution of 1
arc-minute or less. The purpose of operating a high-resolution,
white-light coronagraph in an Earth orbit is to obtain a nearly
continuous patrol of the outer corona over a period of several
months. High-resolution observations from an orbital corona-
83
 SOLAR PHYSICS

graph would provide, for the first time, details of the changes OC-
curring in the outer corona, which may be presumed to ai-
company many solar surface and radio events.
Spectroheliographs in Ultraviolet and Hydrogen Lyman-Alpha
The objectives of the spectroheliograph experiment ( Purcell,
Tousey, and Friedman, NRL) are to carry out spectrohelio-
graphic observations with high angular and wavelength resolu-
tion with the hydrogen Lyman-alpha line, and high angular reso-
lution spectroheliographic observations using the entire emission
line, for certain chromospheric and coronal lines. Only in the
extreme ultraviolet can spectroheliographic observations be made
with lines originating in the high chromosphere and corona.
Observations from an orbiting observatory are required to follow
into the high chromosphere and corona, the morphology and
dynamic processes of solar events originating low in the Sun’s
atmosphere.

300 A-1300 A Scanning Spectroheliometer

The proposed spectroheliometer (Goldberg, Reeves, and Park-
inson, HCO) will greatly improve the spatial resolution of the
solar disk from about 1 arc-minute to 5 arc-seconds. The im-
proved resolution should be good enough for studies of the fine
structure of active regions and, possibly, of the nonhomogeneous
structure of the chromosphere and corona. Spectroheliograms
will also be obtained of areas 5 arc-minutes square with a spatial
resolution of 5 arc-seconds at any desired wavelength in the
300-A-1 300-A range.
Future flights should provide considerably more data on the
distribution of ultraviolet and X-ray sources across the Sun and
their correlation with visible and radio sources. The relatively
sparse data on line spectra in the ultraviolet and X-ray region
will be augmented by both rocket and satellite flights. The
numerous X-ray monitors should enable flares and other dis-
turbances to be observed in greater numbers and in more detail,
while the continuation of measurements will extend through a
maximum of the solar cycle.

84
 THE FUTURE PLIGHT PROGRAM

FAR DISTANT FUTURE POSSIBILITIES

When technological advances permit such missions, a close

vantage point for solar observation will naturally allow higher
angular resolution than the largest feasible Earth-bound or Earth-
orbiting telescopes. The technological areas requiring develop-
ment are upper stage vehicle power and flight control; wide band-
width communication (data and control telemetry) over the
large distances involved ;thermal protection for experiments and
spacecraft subsystems; and high output power supplies to sup-
port the communication system.
Early feasibility studies have revealed the scope of engineering
effort required to perform a minimum justifiable optical solar
probe mission. The approximate resolution limit of ground
solar telescopes is '/3 arc-second at 5000 A. In the foreseeable
future, an upper limit to Earth-orbiting solar telescope resolu-
tion might be 1/40 arc-second ( 100-inch diffraction-limited ob-
jective at 2500 A ) . Such a large-aperture telescope certainly
will require manned supporting operations. At 0.2 AU from the
Sun, a 20-inch telescope could achieve comparable angular reso-
lution, and an instrument this size could conceivably operate on
an unmanned mission. Despite its higher capital costs, a manned
100-inch telescope will offer a greatly extended lifetime of re-
search utility compared to the few weeks operation of a probe
near perihelion. On the other hand, the technological develop-
ment required by a 100-inch manned telescope may come much
later than needed by the solar probe.
Close-look telescopic observation is not the only special ad-
vantage of the probe. It would also permit back-of-Sun ob-
servations ; however, these promise little new scientific knowl-
edge. Moreover, it is not clear that such observations done on a
patrol basis would materially improve the ability to forecast
solar flares. The probe would also provide a unique opportunity
for examining a single prominence simultaneously from two
points of view-from Earth and from the probe. However, this
feature is not a compelling reason to perform a probe experi-
ment since solar rotation provides the essential information in a
statistical way from many prominences.
 SOLAR PHYSICS

I n theory, observations closer than 1 AU to the Sun provide a

gain in flux through a given telescope objective. Thus, the probe
should increase signal-to-noise ratio, and in critical cases it might
permit otherwise impossible observations, such as searching for
some faint ultraviolet radiation expected during only one brief
phase of a solar flare. A similar advantage relates to solar
neutron studies from a probe : the average transit time through 1
AU of neutrons is comparable to their half-life. Therefore, the
close-in probe should encounter a much greater increase of
neutron flux than the inverse square law would indicate.
After reviewing the advantages and shortcomings of the close-
in solar probe, the majority of solar astronomers have recom-
mended that its application should be in the area of local measure-
ments of the solar plasma. The energy spectra and flux of solar
protons and M-region particle streams and the strength and di-
rection of the associated magnetic fields can be discovered only
by in situ measurements. The morphology of cosmic-ray modu-
lation phenomena, and particularly the character of the inter-
planetary storm, continue to rank high in interest as space re-
search objectives.

86
 chapter 10

Summary and Conclusions

B Y THE BEGINNING of the 6-year period, 1958 to 19M, the

first attempts to examine the solar soft X-ray and ultra-
violet spectrum had been made. During this period these meas-
urements have been refined so that the whole spectrum is now
known fairly accurately under quiet-Sun conditions. Measure-
ments have also demonstrated the large variations in X-ray flux
that occur during solar disturbances and the relation between
these and ionospheric phenomena. Thus far, however, only the
simpler aspects of these phenomena are understood, and a con-
siderable extension of measurement techniques is needed to
understand the numerous solar processes. Similarly, high wave-
length and angular resolution have only recently been obtained
in solar soft X-ray measurements, and much remains to be done.
Some unresolved solar problems are briefly discussed below.
THERMODYNAMIC STRUCKJRE

Great effort in solar physics is directed toward a knowledge

of the thermodynamic parameters that govern the solar proper-
ties in both quiet and disturbed conditions. The simplifying
assumptions of energy equilibrium established by radiative proc-
esses only, hydrostatic equilibrium and local thermodynamic
equilibrium are not valid and there is considerable controversy
over the magnitudes, and sometimes the direction, of departures
from these assumptions. An example is the extremely rapid rise
in electron temperature above the photosphere from about
5 x lo5OK at the photospheric surface to about 10"O K at altitudes
above 10' kilometers. One of the crucial problems of the solar
atmosphere is the nature of this nonradiative energy source and
its possible connection with subphotospheric convection and the
presence of wave disturbances.

209-916 -7 87
 SOLAR PHYSICS

GEOMETRIC STRUCTURE

Departures from equilibrium conditions are indicated by the

presence of granulation, faculae and sunspots in the photosphere,
together with a mottled appearance of the lower chromosphere.
In addition, the upper chromosphere above about 3000 kilo-
meters consists entirely of geyserlike spicules having a short
lifetime. An accurate description of the detailed structure
throughout the whole atmosphere presents another problem.
VELOCITY FIELDS

Mass movements in the solar photosphere and chromosphere

have been detected by observations of the Doppler shift of spec-
tral lines, but such results allow extremely wide ranges of inter-
pretation. The energy transported in mass movements is ex-
tremely large, thus representing an important field for future
study.
FLARES

The study of flares involves all the difficulties in understanding

the quiet Sun together with the added complexity that flares are
a time-dependent phenomena and are related to the production
of large numbers of high-energy particles. Flare phenomena
present a considerable challenge since they involve the whole
electromagnetic spectrum down to hard X-rays and the particle
spectrum from a few hundred eV to cosmic-ray energies.
CORONAL PHENOMENA

All observations show large-scale structure in the corona.

Spectroscopic results, however, indicate additional small-scale
structure. Some of these results might be considered further in
connection with the study of heating mechanisms and of the deg-
radation of mechanical energy. The type of analysis used for
the transition-region ultraviolet spectrum could be extended to
include X-ray, visible, and radio spectra. I t would be desirable
to have observations covering as wide a range of wavelengths as
can be made at the same time.
Other problems are posed by the apparent differences in com-
position between the corona and the photosphere. While the
88
 SUMMARY AND CONCLUSZONS

assumption of some departure from spherical symmetry is prob-

ably needed to interpret these data, this has not yet been con-
clusively demonstrated. At present not enough information is
available to permit postulation of a two or more component
atmosphere for the transition region. Further, accuracy must
be improved, especially for white-light observations, which are
important in determining the electron density.
SOLAR MAGNETIC FIELDS

Polarization of radio waves and the trajectories of some coronal

and chromospheric features show that magnetic fields are present
near the Sun. It would be of interest to know the field configura-
tions in detail near the Sun and also further out. For this pur-
pose, solar probes and magnetometer techniques would provide
useful information.
FURTHER UNSOLVED PROBLEMS

Thus far in the measurement of solar soft X-ray and ultra-

violet flux,the time-resolution limit of transient phenomena such
as flares has been set by the detection instruments, and it would
be desirable to improve resolution to better than 1 second. Such
time resolution would give further information on the processes
occurring during solar disturbances and on the volumes of
material participating in the radiation emission.
Earthbound monitoring of solar radiofrequency radiation be-
low 10 Mc/sec is limited by ionospheric absorption and reflec-
tion, but such monitoring has not yet been attempted from above
the atmosphere.

89
 1. MAYER, U. : Space Sci. Rev., vol. 3, 1964, p. 781.
2. ERGIN,E. I.: Progress in Astronautics and Astronornics. Vol. XIII.
Academic Press, 1964, p. 3.
3. DOLDER, F. P.; BARTAE, 0. E.; MERCURE, R. C., JR.; GABLEHOUSE, R.
H.; AND LINDSAY,J. C.: Space Research I11 (W. Priester, ed.),
North-Holland Pub. Co. (Amsterdam), 1963, p. 1207.
4. BURNIGHT, T. R.: Phys. Rev., ser. 2, vol. 76, 1949, p. 165.
5. BLAKE,R. L.; CHUBB,T. A.; FRIEDMAN, M.; AND UNZICKER, A. E.:
Astrophys. J.,vol. 137,1963, p. 3.
6. POUNDS, K. A.; AND BOWEN,P. J.: Monthly Notes, Roy. Astron. Soc.,
vol. 123,1962, p. 348.
7. POUNDS,K. A.: Ph. D. Thesis. University College, London, 1961.
8. ATKINSON,P. A.; AND POUNDS, K. A.: J. Phot. Sci., 1964.
9. KREPLIN,R. W. : Ann. Geophys., vol. 17, 1961, p. 19.
10. KREPLIN, R. W.; CHUBB,T. A.; AND FRIEDMAN, H. : J. Geophys. Res.,
vol. 67, 1962, p. 2231.
1 1 . ACTON,L. W. ; CHUBB, T. A. ; KREPLIN,R. W.; AND MEEKINS, J. F.:
J. Geophys. Res., vol. 68,1963, p. 3335.
12. MANDEL’STAM, S. L.; TINDO,I. P.; VORON’KO, Yu. K.; VASIL’YEV,
B. N.; AND SURGYIN, A. I.: Planetary Space Sci., vol. 11, 1963, p. 61.
13. NEWELL,H. E.: High Altitude Rocket Research. Academic Press,
1953.
14. POUNDS, K. A. ;AND WILLMORE, A. P. : Space Research I11 (W. Priester,
ed.) ,North-Holland Pub. Co. (Amsterdam), 1963, p. 1195.
15. CULHANE, J. L.; POUNDS, K. A.; SANFORD, P. W.; AND WILLMORE, A.
P.: Space Research IV (P. Muller, ed.), North-Holland Pub. Co.
(Amsterdam), 1964, p. 741.
16. LUKIRSKII,A. P.; RUMSH,M. A.; AND SMIRNOV, L. A.: Opt. Spectr.
(U.S.S.R.), English Transl., vol. 9, 1960; no. 4, p. 265; no. 5, p. 343.
17. YEFREMOV, A. I.; PODMOSHENSKN, A. L. ; IVANOV, M. A.; NIKIFOROV,
V. N.; AND YEFIMOV, 0. N.: Artificial Earth Satellites, vol. 10, 1961,
p. 48.
18. RENTSCHLER, H.; HENRY,D.; AND SMITH,K. 0.:Rev. Sci. Instr.,
vol. 3,1932, p. 794.

91
 SOLAR PHYSICS

19. DUNKLEMAN, LAWRENCE : Ultraviolet Photodetectors. NASA T N

D-1718,1963.
20. TAFT,E.; AND APKER,L.: J. Opt. SOC.Am., vol. 43, 1953, p. 81.
21. GOODRICH, G. W.; A N D WILEY,W. C.: Experiments with the Bendk
Continuous-Channel Multiplier. Image Intensifier Sym. NASA I

SP-2,1962, pp. 211-213. I

22. HUNTER, W. R. : Space Research I11 (W. Priester, ed.) ,North-Holland

I
Pub. Co. (Amsterdam), 1963,p. 1187.
23. GOODRICH, G. W.; AND WILEY,W. C.: Rev. Sci. Instr., vol. 32, 1961,
p. 846.
24. STERNGLASS, E. J.; A N D FEIBLEMAN, W. A.: U.S. Patent No. 2,898,499, ‘
1961.
25. CHUBB,T. A.; FRIEDMAN, H.; KREPLIN,R. W.; BLAKE,R. L.; AND
UNZICKER, A. E.: Mem. Soc. Roy. Sci. Liege, Ser. 5, vol. 4, 1961,
p. 228.
26. BAEZ,A. W. : J. Opt. SOC.Am., vol. 51, 1961, p. 405.
27. MOLLENSTEDT, G.; VOYGROTE,K. H.; A N D JONSSON, C.: Paper pre-
sented at Third Intern. Symp. X-Ray Optics and X-Ray Micro-
analysis (Stanford), 1962 (Pattee, Gosslett, and Engstrom, eds.) ,
Academic Press, 1963, p. 73.
28. HENKE,B. L.; A N D MUELLER, W. M.: Advances in X-Ray Analysis.
Vol. IV. Plenum Press (New York), 1961, p. 244.
29. GIACCONI, R.; GURSKY, H.; PAOLINI,F. R.; AND ROSSI,B. R.: Phys. 1
Rev. Letters, vol. 9, 1962, p. 439.
30. GIACCONI, R.; HARMON, N. F.; LACEY,R. F.; AND SZILAGYI, Z.: An
X-Ray Telescope. NASA CR-41, 1965.
3 1. GRAY,S. : Electronic UV Imaging Techniques. Proc. Natl. IAS-ARS
Joint Meeting (Calif.), 1961.
32. BYRAM, E. T.; CHUBB, T.; A N D FRIEDMAN, H.: Rocket Exploration of
the Upper Atmosphere (R. L. F. Boyd, M. Seaton, and H. s. w.
Massey, eds.) ,Pergamon Press, Ltd. (New York), 1954.
33. MANDEL’STAM, C. L.; TINDO,I. P.; VORON’KO, Yu. K.; SURGYIN, A. I.;
A N D VASIL’YEV, B. N.: Planetary Space Sci., vol. 9, 1962, p. 977. I
34. SAWYER, G. A.; JAHODA, F. C.; RIBE,F. L.; A N D STRATTON, T . F.: J.
Quant. Spectr. Radiative Transfer, vol. 2, 1962, p. 571.
35. NEUPERT, W. M.; BEHRING, W. E.; WHITE,W. A.; A N D LINDSAY, J. c . : I
X-614-62-156, Goddard Space Flight Center, 1962.
36. TOUSEY, R.: Space Sci. Rev., vol. 2, 1963, p. 3. ,I
37. PURCELL, J. D.; PACKER, D. M.; A N D TOUSEY,R.: Space Research.
North-Holland Pub. Co. (Amsterdam), 1960, pp. 581-589 and pp.
594-598.
38. DETWILER, C. R.; PURCELL, J. D.; A N D TOUSEY, R.: Mem. Soc. Roy.
Sci. Liege, Ser. 5, vol. 4, 1961, p. 254.
39. RENSE,W. A , : Phys. Rev., vol. 91, 1953, p. 299.

92
 REPERENCES

40. VIOLETT,T.; AND RENSE,W. A.: Astrophys. J., vol. 130, 1959, p. 954.
41. AUSTIN, W. E.;PURCELL, J. D.; AND TOUSEY, R.: Astron J., vol. 67,
1962, p. 110.
42. HINTEREGGER, H. E.: Telemetering Monochromator Measurements of
Extreme Ultra-violet Radiation. Space Astrophysics (W. Liller, ed.),
McGmw-Hill, 1961, pp. 34-95.
43. PURCELL, J. D.; AND TOUSEY, R.: Mem. Soc. Roy. Sci. Liege, Ser. 5,
vol. 4, 1961, p. 274.
44. DETWILER,C. R.; AND PURCELL, J. D.: J. Opt. Soc. Am., vol. 52, 1962,
p. 597.
45. TOUSEY, R.:Quart. J. Roy. Astron. Soc.,vol. 5,1964,~.123.
46. ALLEN,C. W. : Space Sci. Rev., vol. 4,1965, p. 91.
47. HINTEREGGER, H. E.: Rocket Spectra of the Chromosphere. Paper
presented at Symp. on the Solar Spectrum (Utrecht), Aug. 1963.
48. HALL,L. A.; DAMON, K. R.; AND HINTEREGGER, H. E. : Space Research
I11 (W. Priester, ed.), North-Holland Pub. Co. (Amsterdam), 1963,
pp. 745-77 1.
49. PURCELL, J. D. ; GARRETT, D. L. ;AND TOUSEY, R. : Space Research I11
(W. Priester, ed.), North-Holland Pub. Co. (Amsterdam), 1963,
p. 781.
50. HINTEREGGER, H. E.; HALL,L. A.; AND SCHWEITZER, W.: Astrophys. J.,
vol. 140,1964, p. 319.
51. ZIRIN,H.: Astrophys. J., vol. 140, 1964, p. 1332.
52. PO-ITASCH, S.R. : Space Sci. Rev., vol. 3,1964, p. 816.
53. HOUSE, L. L.: Astrophys. J. Suppl. No. 81, vol. 8,1964,~.307.
54. GOLDBERG, L.;MULLER, E. A.; AND ALLER,L. H.: Astrophys. J. Suppl.,
vol. 5, no. 45,1960, p. 1.
55. MORTON, D. C.; AND WIDING, K. D.: Astrophys. J., vol. 133,1961, p. 596.
56. ZIRIN,H.; AND DIETZ,R. D.: Astrophys. J., vol. 138, 1963, p. 664.
57. POUNDS, K. A.; et al.: Proc. Roy. SOC.London, Ser. A, vol. 281, 1964,
p. 538.
58. HULBERT, E. 0.:Phys. Rev., vol. 53,1938, p. 344.
59. VEGARD, L. : Geophys. Publ., vol. 12, 1938, p. 5.
60. EDLEN, B.: Z. Astrophysics, vol. 22,1942, p. 30.
61. HOYLE,F.; AND BATES,D. R.: Terrest. Magnetism and Atmospheric
Elec., vol. 53, 1948, p, 51.
62. FRIEDMAN, H.; LICHTMAN, S . W.; AND BYRAM, E. T.: Phys. Rev., VO!.
83, 1951, p. 1025.
63. FRIEDMAN, H. : Rept. Progr. Phys., vol. 25,1962, p. 163.
64. ELWERT, G.: J. Geophys. Res., vol. 66,1961, p. 391.
65. STEIN,W.A.; AND NEY, E. P.: J. Geophys. Res., V O ~ . 68, 1 x 3 , p. 65.
66. SCHKLOVSKY: Nature, vol. 202,1964, p. 275.
67. ACTON,L. W.:X-Radiation of the Sun. Thesis, University of c o b
rado, 1964.

93
 SOLAR PHYSICS

68. FRIEDMAN, H.; AND CHUBB,T. A.: Rept. of Conf. on Physics of the I

Ionosphere (Cambridge), 1954. Phys. SOC.(London), 1955, p. 58.

69. BLAKE,R. L.; CHUBB,T. A.; FRIEDMAN, H.; A N D UNZICKER, A. E.:
I
Science, vol. 146, 1964, pp. 1037-38.
70. PETERSON, L. E.; A N D WINCKLER, J. R.: J. Geophys. Res., vol. 64,1959, ,
p. 697. I

71. CHUBB, T. A. ; FRIEDMAN, H.; AND KREPLIN, R. W. : J. Geophys. Res.,

vol. 65, 1960, p. 1831.
72. ANDERSON,K. A.; AND WINCKLER, J. R.: J. Geophys. Res., vol. 67, 4
1962, p. 4103.
73. SEATON, M. J.: 1964, Planetary Space Sci., vol. 12, 1964, p. 55.
74. DE JAGER,C.; A N D KUPERUS, M.: Bull. Astron. Inst. Neth., vol. 16,
1961, p. 71.
75. LINDSAY, J. C. : AIAA Preprint 63-470, AIAAf CAS1f RAeS 9th Anglo-
American Conference, 1963.
76. WHITE,W. A. : Space Research IV (P. Muller, ed.) , North-Holland
Pub. Co. (Amsterdam), 1964,p. 771.
77. WHITE,W. A.: Planetary Space Sci., vol. 12, 1964, p. 379.
78. FROST,K. J. : Paper presented at COSPAR Symposium, May 1963.
79. NEUPERT, W. M.; BEHRING, W. E.; AND LINDSAY, J. C.: Space Research
IV (P. Muller, ed.) , North-Holland Pub. Co. (Amsterdam), 1964,
p. 719.
80. NEUPERT, W. M. : Space Research IV (P. Muller, ed.) , North-Holland
Pub. Co. (Amsterdam), 1964, p. 731.
I
81. HALLAM, KENNETH L.: Solar Flares in the Light of Hydrogen Lyman-
Alpha. AAS-NASA Symposium on the Physics of Solar Flares.
NASA SP-50, 1964, pp. 63-64.
Articles of General Interest

HINTEREGGER, H. : Rocket Spectra of the Chromosphere. Paper presented

at Symp. on the Solar Spectrum (Utrecht), 1963.
TOUSEY, R.: Extreme Ultraviolet Spectrum of the Sun. Space Sci. Rev.,
vol. 11, 1963, p. 3.
Report of Comm. 44: Observations au de hors 1’Atmosphere Terrestre.
IAU 12th General Assembly (Hamburg), 1964.
VANDE HULST,H. C.; DE JAGER, C.; A N D MOORE,A. F., EDS.: Space Re-
search 11. North-Holland Pub. Co. (Amsterdam), 1961.
PRIESTER, W., ED. : Space Research 111. North-Holland Pub. Co. (Amster-
dam), 1963.
ALLEN,C. W.: The Interprrtation of the XUV Solar Spectrum. Space
Sci. Rev., 1964.
FRIEDMAN, H.: Ultraviolet and X-rays From the Sun. Ann. Rev. Astron.
Astrophys., vol. 1, 1963, p. 59.

94
 REPERENCES

IAU Symposium No. 23 : Astronomical Observation From Space Vehicles,

‘Liege, 1964.
GOLDBERG, L. : Space Astronomy. P m . IAU 12th General Assembly, Ham-
burg, 1964.
TOUSEY, R.: George Darwin Lecture. Quart. J. Roy. Astron. Soc., vol. 5,
1963, p. 123.
BOYD,R. L. F.: Techniques for the Measurement of Extra-Terrestrial Soft
X-Radiation. Space Sci. Rev., vol. 4, 1965, p. 35.
FRIEDMAN, H.: Solar Observations Obtained From Vertical Sounding.
Rept. Progr. Phys.,vol. 25, 1962, p. 163.
SMITH,H. J. ;AND SMITH,E. : Solar Flares. Macmillan Co., 1963.
LINDSAY, J. C.: The Solar Extreme Ultraviolet Radiation (1-400 A).
Planetary Space Sci., vol. 12,1964, p. 379.

95
US. GOVERNMENT PRINTING OFFICE: IS66 0-2041)18
 Y

“The aeronautical and space activities of the United States shall be

conducted so as to contribute . . . t o the expansion of human Rnowl-
edge of phenomena in the atmosphere and space. The Adrninistrdion
shall provide for the widest practicable and appropride dissemination
uf infurmation concerning its activities and tbe results thereof.”
-NATIONAL AERONAUTICS AND SPACE ACr OF 1958

NASA SCIENTIFIC AND TECHNICAL PUBLICATIONS

TECHNICAL REPORTS: Scientitic and technical information considered

important,complete, and a lasting contribution to existing knowledge.
TECHNICAL NOTES. Information less broad in scope but nevertheless of
importance as a contribution to existing knowledge.

TECHNICAL MEMORANDUMS Information receiving limited distribu-

tion because of preliminary data, security classification, or other repsons.

CONTRACTOR REPORTS Technical information generated in connection

with a NMA contract or grant and released undu NASA auspices.

TECHNICAL TUNSLATIONS Information published in a f o d p

language considered to merit NASA distribution in English.
SPECIAL PUBLICATIONS Information derived from or of value to NASA
activities. F’ublications include conference proceedings, monographs, data
compilations, handbooks, soumbookr, and s p e d bibliographies.

TECHNOLOGY UTILIZATION PUBLICATIONS: Information on tech-

nology used by NASA that may be of pnrticulpr interest in commercial and other
nonaerospe applications. Publications include Tech Briefs; Technology
Utilization Reprta and Notes; and Technology Surveys.

h t a i k on the availability of these publications may be obtained from:

SCIENTIFIC AND TECHNICAL INFORMATION DlVlSlON

NAT I0 NA L A E R 0NA UT I C S A N D S P A C E A D M I NI ST R AT I 0 N
Washington, D.C. PO546