...molecule

Nowadays, it is clear that measurements in the infrared are useful for registering the presence of excited H₂.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...users

A second orbital flight of IMAPS will have time set aside for guest observers -- see §8.1.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...fortunately

We say ``fortunately'' because, as things turned out, the instruments that were selected had their first flight on the Astro-1 mission that occurred some 6 years after the first flight of IMAPS on a sounding rocket.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...missions

A Spartan carrier is a spacecraft that has support systems nearly identical to those of a sounding rocket (gas jet attitude control system, power from batteries, etc.), but is carried into orbit by the Shuttle and deployed to fly on its own. After observing, the Shuttle retrieves the Spartan and returns it to the ground.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...Astro-SPAS

Astro-SPAS is a spacecraft that is carried into orbit by the Shuttle and is very similar to Spartan, only much larger. Astro-SPAS was designed and built in Germany. The Astro-SPAS mission slated to carry ORFEUS was called ORFEUS-SPAS.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...ORFEUS

ORFEUS stands for Orbiting Retrievable Far and Extreme Ultraviolet Spectrometers. It consists of a 1 meter diameter telescope that feeds two spectrographs, one built by the Astronomisches Institut Tübingen and another built at the University of California at Berkeley.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...Lomb

The division that makes gratings is now operated by the Milton Roy Corp.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...section.

As far as we are aware, the only other space experiment that uses a detector that operates on the same principle as ours is one developed by Carruthers & Seeley (1992).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...current.

After removing the fixed pattern caused by variations in dark current from one pixel to the next (§7.3) (but stable over short periods of time), we find a random noise equivalent to 170 e^-pixel^-1. Most of this noise is from the on-chip amplifier of the CCD. Statistical fluctuations in the CCD's room-temperature dark current of around 20,000 e^-pixel^-1s^-1 are responsible for only 50 e^-pixel^-1 rms noise in any single frame with an exposure time of 0.067 s.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...${\rm pixel}^{-1}$.

This never happened during the first flight on the ORFEUS-SPAS satellite discussed in §§6-6.2.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...images

Throughout this paper, we employ the term ``frame'' to mean a single readout of the CCD, while the words ``image'' or ``exposure'' refer to many such frames added together.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...seconds.

Strictly speaking, only 255 frames should be added if one wants to be sure an overflow does not occur. When this happens for a very bright source, no information is lost since the outcome is modulo 2¹⁶.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1000Å.

This effective area was achieved on the sounding rocket flights discussed in §2 but not on the ORFEUS-SPAS flight, owing to the damage to the detector's photocathode (§6.1.3).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...picture

With discrete Fourier transforms, a discontinuity between the top/bottom and left/right edges in the original picture produces an unwanted, very strong signal. This interference was eliminated by artificially forcing these edges to join each other over a smooth transition.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...horizontal

The IDL WARP_TRI routine was invoked for this purpose.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...directly

i.e., through a literal application of the IDL WARP_TRI routine to the raw measurements.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...evaluated

Through the use of the IDL POLY_WARP routine.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...exposure.

If the order cuts were strictly sinusoidal, it would be impossible to differentiate between vertical motion and just a weaker exposure above a larger background level, since either effect would attenuate the primary frequency.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...separations.

We are not aware of any underlying principle that explains this phenomenon, but we are grateful for its existence.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...picture

This was done by invoking a simple linear transformation for the coefficients in the equations that convert the irregular mapping of the detector into true geometrical coordinates, as outlined in §7.6.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...format)

b was determined by examining the shapes of bright orders next to ones where there was total absorption, such as a star's bright P-Cygni emission peak from O VI right above the Lyman-$\beta$absorption line.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...20.

This calculation assumes that the guest observing program has the same 1/5 fraction of the observing time devoted to IMAPS, just as for the principal investigator teams. The actual ratio will depend on a competitive selection of proposals that should be made on the basis of scientific merit and technical feasibility, without regard to which instrument is being used.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.