B. Haisch, S. Bowyer, R.F. Malina
                    Center for EUV Astrophysics,
            2150 Kittredge St., University of California,
                   Berkeley, California 94720, USA


   The successful launch of the NASA Extreme Ultraviolet Explorer on June 7,
1992, is the culmination of nearly 30 years of effort at the University of
California at Berkeley to open up the field of extreme ultraviolet astronomy.
We present a brief introduction to this field and an overview of the satellite
instrumentation to provide the context for this set of articles discussing the
design and operation of the mission and the analysis of the data.


   Surely one of the world's long lists would be the litany of impossibilities
that came to pass. The field of extreme ultraviolet (EUV) astronomy thus joins
the ranks of the automobile, the airplane and space travel as once-infamous
concepts transformed into commonplace reality. It was long believed that the
interstellar medium (ISM), assumed to be cold, would be entirely opaque at EUV
wavelengths (shortward of the 912 A hydrogen Lyman absorption edge down to soft
X-rays) even to the nearest star system (Alpha Centauri A and B along with
Proxima Centauri), a mere 1.3 parsecs away [1]. So when NASA in the vigor of
its youth reached for space horizons to explore in the 1960s, nowhere in the
flurry of reports was space astronomy in the extreme ultraviolet to be found.
   The presence of a hot, million degree gas within the galaxy had been pre-
dicted by Spitzer [2], and this was detected by Bowyer, Field and Mack [3].
Moreover, the discovery of the patchy nature of the ISM determined from 21-cm
maps suggested that the dense (1 hydrogen atom cm^(-3)), cold phase that is
so opaque to the EUV might be confined to interstellar clouds, in pressure
equilibrium with a hotter, hence much less dense and also ionized state of
far great- er EUV transparency.  Believing that this could well modify the
picture of a hopelessly opaque ISM, Bowyer's maverick group of researchers at
the University of California in Berkeley examined and reexamined the opacity
of the ISM until finally in 1974 in an Astrophysical Journal article, whose
abstract was followed by a quote from Milton's Paradise Lost (definitely not
a customary practice), they were able to make a credible prediction that the
ISM might be far less opaque than had been nearly universally assumed [4].
   Encouraged by this, the Berkeley sounding rocket experiments began to search
for astronomical EUV sources, and a major NASA mission was actually selected
for flight: the OSO-J satellite. As luck would have it, the OSO series was
terminated after the flight of OSO-I!  Foiled by the alphabet but not being
accustomed to throwing in the towel, the Bowyer group seized the opportunity
to build and fly a small EUV telescope on the 1975 Apollo-Soyuz mission. Dur-
ing 20 hours of pointing by the crew, 30 preselected stars were searched for
evidence of EUV emission.  This paid off with the detection of radiation from
two white dwarf stars [5,6], the dwarf nova SS Cygni [7], a flare on Proxima
Centauri [8], and emission from the ISM itself [9].  That same year NASA re-
leased a call for Explorer-class projects, to which the group responded with
a proposal for a mission to do an all-sky survey.  This was selected as a new
NASA mission in 1976 and scheduled for flight in 1982. In fact it would be ten
years later than that before the Extreme Ultraviolet Explorer at last trained
its telescopes on the EUV sky.


   The Extreme Ultraviolet Explorer satellite launched on June 7, 1992, into
a near-Earth (550 km) orbit. Following seven weeks of in-orbit checkout, the
science mission began on July 24.  The EUVE mission now underway has three
distinct science objectives.

 (1) To carry out an all-sky survey in the entire extreme ultraviolet in four
	passbands using three scanning telescopes covering the 50-740 A wave-
	length range.

 (2) To survey simultaneously a 2x180 degree swath of sky along the ecliptic
	in two passbands (65-360 A), probing to much deeper sensitivity levels
	(factor of 20) than the scanning telescopes.

 (3) To follow up on the survey with medium-resolution spectroscopy (0.5-2 A)
	of individual detected sources over the 70-760 A range.

   The first two objectives, the all-sky survey and the deep survey, consti-
tute the UC Berkeley principal investigator science program.
   Until as recently as 1990, only two dozen EUV sources had been detected.
That year saw the launching of ROSAT, which carried a primary soft X-ray tele-
scope (SXRT) from the Max-Planck-Institut in Germany [10], and a secondary
instrument provided by the U.K., the Wide Field Camera (WFC), which operates
in the border regime between the EUV and soft X-rays (60 to 200 A, which by
coincidence is also 60 to 200 eV) [11]. These two instruments carried out an
all-sky survey in 1990. However, while some 60000 sources are ultimately ex-
pected to result from the SXRT observations (which are complete but are still
being analyzed), to date only 383 sources have been published in the WFC source
catalogue [12]. 
   The full sky has now been surveyed across the entire EUV spectrum by EUVE.
The survey phase of our mission was completed on January 21, 1993.  Preliminary
results indicate that as many as 2000 sources may be detected once all the data
are processed. This processing is a sensitive and time-consuming procedure be-
cause of the conflicting requirements of extracting the very faintest sources
from a time- and space-variable background while minimizing the number of spu-
rious sources, i.e. background fluctuations mistaken for sources. This process
is described in some detail in the article in this issue by Lewis.
   The EUVE deep survey succeeded in finding sources a factor of 10 to 20
fainter by concentrating entirely on a 2x180 degree swath of sky (while the
scanners swept across the entire sky) and by carrying out these observations
looking down the Earth's shadow cone, thus minimizing the background.  The
object of this experiment was to search for other types of sources which may
not happen to be bright enough to be represented in the all-sky survey.
   The second phase of the mission, consisting of spectroscopy of individual
sources, began immediately after the all-sky survey. Unlike the survey, the
spectroscopy is carried out by NASA-selected guest observers (GOs) whose sci-
ence programs were evaluated and ranked by a review committee. Pointings on
individual targets typically accumulate 40000 s to over 100000 s of data.
Since these are taken during orbital nighttime to minimize background, the
average duty cycle is about 30 to 40 percent; thus these observations may
stretch over several days' time. Some 130 targets were selected for the first
year of GO observations.


   EUVE consists of three modules: the spacecraft, the science payload, and
an interfacing "platform equipment deck" connecting the two and also serving
as the mount for the solar panels.  The EUVE science payload and its relation
to the spacecraft and the platform equipment deck is shown in fig. 1.
   The science payload uses grazing-incidence Wolter-Schwarzschild-type tele-
scopes. Different variations of this optical system are used for the three
different instruments: the three all-sky survey telescopes, the deep survey
imager, and the three spectrometers (see [13] for details). The scanning tele-
scope system is depicted in fig. 2.  The mirrors are made of aluminum cut to
a precise figure by a diamond-turning lathe; they are nickel-plated, polished
and finally gold-coated for maximum reflectivity. Graze angles of 5-10 degrees
are used, which are still very reflective in the EUV but are large enough to
eliminate unwanted X-ray reflections.
   Photon-counting, position-sensitive EUV detectors (invented by M. Lampton
and co-workers and developed for spaceflight by O. Siegmund and co-workers,
all at UC Berkeley) are used. A photon striking the front surface of this high
voltage device results in a cascading cloud of 20 to 30 million electrons
zig-zagging down stacked microchannel tunnels; at the ends of the tunnels,
the position of the charge-cloud is electronically determined, yielding the
original photon position in the field of view.  Such detectors are ideal for
this type of mission: they are stable, rugged, linear in response and solar-
blind, and they provide effectively 1680x1680 independent resolution elements.
   The "multimission modular spacecraft (MMS)" is a 3-axis stabilized satel-
lite with a stellar reference control system that includes two star trackers
and a fine sun sensor. This system allows the satellite to be pointed and to
maintain a position with 60 arcsec accuracy. (During the all-sky survey, the
satellite was operated in a scanning, rather than pointing mode.) Two onboard
tape recorders each have a storage capacity of 103e9 bits. These data are re-
layed at prespecified intervals via a NASA satellite system to the Goddard
Space Flight Center and then to the EUVE Science Operations Center (ESOC) at
the UC Berkeley Center for EUV Astrophysics.  Commands to the satellite and
to the science instruments are transmitted from the ESOC via the same link.

3.1 The Sky Survey Instruments

   Three of the EUVE telescopes are co-aligned and together point at right
angles to the satellite spin axis, as shown in fig. 3. The focal planes of
the three telescopes are divided into quadrants, and four different types of
filters distributed over these quadrants isolate four wavelength passbands
for imaging (see table 1). During the sky survey, the satellite was spun
three times per orbit so as to image a 2 degree wide band of sky in each of
these passbands. Spinning at this rate, rather than once per orbit, was done
to assure that one complete scan could be carried out during satellite night,
to minimize background from scattered solar radiation.  As a source drifted
across the fields of view of the three telescopes, it was detected in the
various filter passbands and its position was subsequently mapped back onto
the sky from satellite pointing data.

3.2 The Deep Survey Instrument

   This instrument is literally half of a telescope: this fourth EUVE telescope
has its mirrors divided into six equal segments. Three of the six segments are
used for spectroscopy (see section 3.3) and the remaining three segments form
a single image onto a field of view that is divided into two halves by differ-
ent filters (see table 1). This telescope is aligned along the spin axis and
during the all-sky survey always pointed in the anti-sun direction.  Aligned
with the axis of rotation, it did not sweep across the sky, but had its field
of view precess along the ecliptic during the six months of the survey. It
thus obtained a much deeper exposure than the scanners within a restricted
2x180 degree swath of sky. This instrument also acts as an EUV photometer dur-
ing spectroscopic observations, an extremely useful function for the many types
of variable sources being observed.

3.3 The Spectrometers

   The other three segments of this telescope each feed light onto a reflection
grating, as shown in fig. 4. This spectrometer design was invented at Berkeley
by M. Hettrick and S. Bowyer. It makes use of variable line-spaced gratings.
This design was predicted to have a number of advantages for grazing-incidence
spectroscopy (see [13,14]), which is indeed proving to be the case for the
EUVE spectrograph.  The three spectrometers operate simultaneously and provide
approximately 1-2 A resolution spectra in the three bands listed in table 1.
All observations following the end of the all-sky survey (apart from the fil-
ling in of gaps) are being made with the EUV spectrometers.


   The main sources of opacity in the EUV are neutral hydrogen and helium. The
absorption cross section for neutral hydrogen peaks at the Lyman edge (912 A)
at a value of approximately s = 6.6e-18 cm^2 per neutral hydrogen atom and
falls off roughly as l^3 for shorter wavelengths, where l is the wavelength.
The attenuation is e^[-s(l) * n_H * d)], where n_H is the density of neutral
hydrogen, d is the distance and a constant hydrogen-to-helium ratio is assumed.
The quantity N_H = n_H * d is called the column density.  It can be visualized
as a cylinder of length  with a base of one square centimeter. If one looks
through the base of the cylinder all the way down its length, one can count
the number of hydrogen atoms seen in projection on the base, which is the col-
umn density, in effect a surface density of hydrogen atoms per square centime-
ter of base area. When the exponent in the attenuation law is greater than five
optical depths, less than 1% of the light will be transmitted, i.e. e^(-t) <
0.01 for t > 5.  In the middle part of the EUV range (l approx 350 A), the
cross section is such (s = 5e-19 cm^2) that a column density of N_H = 1e19
will yield a 1% transmission. In other words a column density of 1e19 is about
as far as one can see at 350 A.  (It will be much less at longer wavelengths
nearer the 912 A Lyman edge.)
   What physical distance does this correspond to? Astronomers used to believe
that the ISM was cold, hence hydrogen would be in the EUV-opaque neutral state
and that the density was a high n_H = 1 atom cm^(-3). If that were the case,
the limiting 1e19 column density would be attained in 1e19 cm, i.e. a mere 3
pc at 350 A and much shorter path lengths at longer wavelengths.
   In fact, it appears that the actual distance over which a 1e19 column den-
sity is attained ranges from tens of parsecs in the plane of the galaxy to
hundreds of parsecs near the north galactic pole [15].  How can this be? There
are two causes. First of all, the overall density appears to be much lower,
and specifically the NEUTRAL hydrogen density is quite low. A popular concept
(not without its detractors) is that we appear to be situated in a hot, mil-
lion-degree bubble of material (which may be quite irregular in shape). To
complicate this picture even further, there does appear to be cool, denser
material in the immediate vicinity of the Sun (say within 3-5 pc), with the
Sun and this surrounding cloud sitting inside the hot, tenuous bubble. Other
cool clouds are thought to be scattered throughout the bubble.
   What is the origin of this bubble, if indeed it is a bubble? Unfortunately
we do not yet know whether it is the result of an explosive event (one or more
more supernovae) which is still expanding and therefore continuing to heat ma-
terial at a shock front interface with the ISM, or whether the bubble is in
stable pressure equilibrium with the external ISM (i.e. a hot, low-density gas
can have the same pressure as a cool dense gas since p is proportional to n*T).
At the risk of mistakenly focusing on the hot theory of the year, it was re-
cently proposed that two riddles could be explained in one fell swoop by iden-
tifying the (until 1992) mystery gamma-ray source, Geminga, with the pulsar
remnant of the very supernova that created the local hot bubble [16]. Tracing
back the proper motion of Geminga and comparing it with the roughly known
topology of the local bubble does suggest a possible origin in this supernova
explosion some 3e5 years ago at a distance from Earth of no more than 60 pc.
It remains to be seen whether this theory will survive the test of time.
   It is expected that EUVE will shed (EUV) light upon the two foremost prob-
lems concerning the local ISM:

 (1) what is the actual structure of the local bubble; and
 (2) how many other cool clouds, like the one surrounding the Sun, are there
	in our bubble?

   The ISM can be probed in the optical spectrum, but only indirectly by seeing
the effect of this intervening material on a background stellar spectrum. It is
like being immersed in a fog bank and having only the pencil beams of distant
lights here and there penetrating the muck to tell us that something is sur-
rounding us. And clearly, in directions where there happen to be no sources
bright enough to yield the high resolution spectra carrying the ISM signature,
nothing can be inferred. On the other hand, EUVE can in principle see the EUV
glow of the interstellar material itself.
   Concerning other cool clouds, it is as difficult to identify these using
sparsely distributed stellar light beams as probes as it would be to map out
the raisins in a coffee cake with a collection of hat pins. But preliminary
indications are that EUVE may be able to see cool clouds as dark shadows in
the EUV sky. The number of clouds sharing our bubble is important because they
are expected to, in effect, evaporate over the course of time, and so the num-
ber of clouds can be related to the age of the bubble by use of statistics,
models and other tricks of the theoreticians' profession.


   A standard procedure in a survey of this kind is to generate an initial list
of bright sources to take stock of the processing algorithms and the types of
sources that are appearing, and to guide the selection of spectroscopy targets.
The all-sky survey Bright Source List contains the types of objects listed in
Table 1 in the paper by Stroozas et al., this issue.  (This list is only the
tip of the iceberg of total sources.)
   Initial results from the deep survey along the ecliptic appear to show about
nine times as many sources per square degree as appear in the all-sky survey.
This is a factor of two fewer than expected, although at this time it is pre-
mature to conclude that this deficiency is significant.
   It is not surprising that the largest categories of objects appearing in
the EUV are late-type stars and white dwarfs; this was expected prior to launch
[17].  Stars of spectral type F through M along the main sequence have X-ray-
emitting coronae, presumably analogous to the Sun, as do evolved stars (giants
and supergiants) of type F and G. On the other hand, K and M giants are not
coronal sources; they lie to the right of the so-called Dividing Line [18]
and indeed appear not to be EUV sources. Based on our knowledge of the Sun,
stellar coronae are believed to consist primarily of hot plasmas at tempera-
tures of one to ten million K confined by magnetic fields into structures known
as coronal loops. Plasmas at these temperatures and coronal densities will emit
both soft X-ray and EUV radiation. Whereas in the soft X-rays this emission is
primarily continuum, in the EUV it is primarily line radiation, which can be
modeled with considerable precision by use of atomic codes developed in the
past 20 years and which can at last be seen in EUV spectra.
   Late-type stars that are at the upper end of EUV and X-ray emission are
called "active stars". This terminology derives from the practice of referring
to spots and surrounding regions of plage on the Sun as "active regions". Since
the days of rocket and Skylab imaging of the Sun in the late 1960s and early
1970s in X-rays, we know that active regions are the areas in which the solar
corona is most highly concentrated in the form of loops. By analogy, it is
thought that stars emit X-rays and EUV radiation from active regions, and the
strong emission of such is now referred to as "stellar activity". The most
active stars are also often observed to flare, as is the Sun. The fact that
the scanning EUVE telescopes sweep across any given star during every orbit
for several days in succession has resulted in several detections of EUV
flares, for example on Prox Cen. (Recall that a Prox Cen flare was one of the
very first EUV detections back in 1975, seen during the Apollo-Soyuz mission;
see [8]). 
   For the known flare star AU Mic, a flare took place during a spectral ex-
posure taken for instrument calibration purposes. Since each photon is time
tagged, the spectrum could easily be separated into a quiescent and a flaring
spectrum, as shown in fig. 5. Individual lines that stand out are being com-
pared to models of emission from plasmas at temperatures in the 10-30 million
K range.
   Unlike for late-type stars, the EUV emission of white dwarfs is blackbody
continuum radiation originating in the photosphere. Wien's radiation law pre-
dicts an inverse relation between peak wavelength and temperature for black-
bodies. Thus, while the photosphere of the Sun (5770 K) peaks at visible wave-
lengths, around 5000 A, white dwarfs with surface temperature between 25000
and 100000 K should emit the bulk of their radiation in the EUV. This is in-
deed the case and explains why white dwarfs are the second most numerous types
of objects in the EUV. But even more interesting than their numbers are the
spectral features apparent in the EUV. White dwarfs, having come to the end
of their thermonuclear evolution, have extremely stable atmospheres that turn
out to be highly stratified by the enormous gravitational fields (in contrast
to the turbulence and convection of the solar atmosphere). This stratification
separates the various elements; however, while on the one hand gravity tends
to make the heavier elements settle far below the observable surface, if the
WD is hot enough, radiation pressure will act to push certain elements of the
proper opacity back up to the surface. Such spectral signatures are most ap-
parent in the EUV, which is truly the prime window for studying WDs. As a
result of effects like this we are finding that the EUV spectra of WDs is
different by as much as a factor of ten from model predictions of only a few
years ago.
   For all the large numbers of late-type stars and white dwarfs, the bright-
est EUV source in the sky turns out to be an entirely different class of ob-
ject, the B star Epsilon Canis Majoris (B2 II).  It had been thought, based
on individual pointed observations prior to surveying the entire sky, that the
hot white dwarf detected during the Apollo-Soyuz mission, HZ 43, would be the
brightest EUV source in the sky. B stars have photospheric temperatures com-
parable to the cooler of the WDs, but since there are fewer B stars than WDs,
the B stars are on average much further away. In fact, CMa is at a distance
of 187 pc. Based on the hot bubble interstellar hydrogen density of 0.005
cm^(-3), we would expect a column density of 3e18 cm^2 along a 187 pc line of
sight. This would attenuate the EUV radiation at 600 A by four optical depths,
i.e.  decrease the radiation by a factor of e^(-4).  The fact that the B-star
photosphere is so bright leads to the conclusion that the hot bubble stretches
at least that far in the direction of Epsilon CMa and that the average hydrogen
density in that part of the sky may be even lower than the average.
   It is now obvious that the ISM in many directions is far more tenuous than
had been expected even by the most ardent EUV enthusiast.  Almost twice as far
away (estimated at 300 pc) as the B star, EUVE has detected time-variable ra-
diation from a magnetic cataclysmic variable.  These "AM Her-type stars", named
after the prototype as is typical in astronomy, consist of a highly magnetized
white dwarf in orbit about a nearby companion from which mass is being drawn
by gravitational tidal forces. Because the magnetic field of the WD prevents
formation of an accretion disk, the material must flow onto the WD at the
poles. Via a process involving shock formation and production of hard X-rays,
EUV radiation is finally formed. Variations in optical light attest to signi-
ficant changes in the accretion rate, which we now are able to observe in the
   If the distribution of cool interstellar material (which is the most opaque
to EUV radiation) outside the bubble is pictured as a distribution of clouds
in our region of the galaxy, then it appears that there are sufficient breaks
in the clouds in some directions to afford us a view of deep space at distances
of megaparsecs. At a distance of 400 million parsecs, the BL Lacertae object
PKS 2155-304 radiates profusely across the spectrum from the visible to hard
X-rays. Moreover, the bulk of the luminosity, which exceeds that of our own
Milky Way galaxy, must be concentrated in a volume the size of our solar sys-
tem, or smaller, given the observed variations in emission on time scales of
hours. EUVE observed this object over a 1.3 day interval and found significant
variations during that time.
   All the observations discussed so far have been of point sources, but EUVE
is capable of resolving extended objects. Given the faintness in general of
EUV sources and the rather significant background "noise" of the sky and the
near-Earth environment at these wavelengths, there are only a handful of ex-
tended objects that will emerge from the fully processed skymaps.  Two such
objects are the Cygnus Loop and the Vela supernova remnants, shown in the
color foldout section of this issue.


   Where does EUV astronomy go from here? The future of the field is more
likely to be determined by political than scientific considerations. Whereas
in former times a successful mission would naturally lead to a more sensitive
follow-on program, the resources are no longer adequate to follow where the
science leads. In times of tight budgets and with a view to satisfying the
requirements of "programmatic balance", the pull is actually in the opposite
direction. A successful mission will be seen as having satisfied the needs of
a given discipline for the foreseeable future.
   EUVE combined the two discrete functions of exploratory survey and follow-
on spectroscopy into a single mission, and as such has been a cost-effective
program in the spirit of the 1990s even though it was conceived in the 1970s
and honed and refined in the 1980s. While there is not liable to be a true
follow-on to the successful EUVE mission in the foreseeable future, there are
at least two upcoming opportunities to expand the EUV horizon via spectroscopy.
   The Orbiting Retrievable Far and Extreme Ultraviolet Spectrometer (ORFEUS)
will be a payload on the Astro-SPAS system to be carried aloft on the Shuttle,
released to do scientific investigations for a number of days, and then re-
trieved and returned to Earth [20]. This is a collaboration involving the
Astronomical Institute at Tubingen, the Landessternwarte Heidelberg, and the
UC Berkeley Center for EUV Astrophysics.
   A major NASA mission scheduled for the year 1999 is the Far Ultraviolet
Spectroscopic Explorer (FUSE-Lyman; see [20]). Its primary mission will be
spectroscopy in the 900--1200 AA region, but the prospects look good that a
shorter-wavelength extreme ultraviolet capability will be included.
   For a former impossibility, the field of EUV astronomy has come a long way.


1. L.H. Aller, Publ. Astr. Soc. Pacific, 71, 324, (1959).
2. L. Spitzer, Jr., ApJ, 124, 20, (1956).
3. S. Bowyer, G. Field, and J. Mack, Nature, 217, 32, (1968).
4. R. Cruddace, F. Paresce, S. Bowyer, and M. Lampton, ApJ, 187, 497, (1974).
5. M. Lampton, B. Morgan, F. Paresce, R. Stern, and S. Bowyer, ApJ Letters,
	203, L71, (1976).
6. B. Margon, M. Lampton, S. Bowyer, S. Stern, and F. Paresce, ApJ Letters,
	210, L79, (1976).
7. B. Margon et al., ApJ, 224, 91, (1978).
8. B.M. Haisch, J.L. Linsky, M. Lampton, F. Paresce, B. Morgan, and R. Stern,
	ApJ Letters, 213, L119, (1977).
9. R.A. Stern and S. Bowyer, ApJ, 230, 755 (1979).
10. J. Trumper, et al., Nature, 349, 579, (1991).
11. M.A. Barstow, and R. Willingale, JBIS, 41, 345, (1988).
12. K.A. Pounds, et al., MNRAS, 260, 77, (1993).
13. S. Bowyer and R. Malina, in Extreme Ultraviolet Astronomy, p. 397, (1991).
14. M.C. Hettrick and S. Bowyer, Applied Optics, 22, 3291m (1983).
15. D.P. Cox and R.J. Reynolds, Ann. Revs. Astr. Ap., 25, 303, (1986).
16. N. Gehrels  Wan Chen, Nature, 361, 706, (1993).
17. S. Bowyer, R.F. Malina, and H. Marshall, JBIS, 41, 357, (1988).
18. B. Haisch, J.H.M.M. Schmitt and A.C. Fabian, Nature, 360, 239 (1992).
19. M. Hurwitz and S. Bowyer, in Extreme Ultraviolet Astronomy, p. 442, (1991).
20. W. Moos and S. D. Friedman, in Extreme Ultraviolet Astronomy, p. 457,


Figure 1:  The three basic modules composing the Extreme Ultraviolet Explorer.
Figure 2:  Cross-sectional schematic of the two Wolter-Schwarzschild survey
	telescopes:  the shorter wavelength (a) and the longer-wavelength (b)
Figure 3:  During the all-sky survey, the spacecraft rotates about the spin
	axis, letting the three survey telescopes sweep out a great circle on
	the sky while the Deep Survey telescope images the ecliptic. As the
	spin axis precesses the entire sky is imaged in six months, and a 2x18
	degree swath along the ecliptic is deeply exposed.
Figure 4:  Schematic illustration of the two grazing-incidence mirrors and the
	variable line-space grating composing one of the three spectrometers.
Figure 5:  The short-wavelength spectrum of the flare star AU Mic showing qui-
	escent coronal emission (light) and a flare (heavy) that took place
	during the exposure.

Page created by
Last modified 10/14/98