So numerous were the interesting findings, that hardly anyone noticed that infrared astronomy had just reached two of its holy grails and was hurtling past them at great speed.

The first of these is the detection, with the Short Wavelength Spectrometer, SWS, of the lowest rotational transition of molecular hydrogen, the S(0) transition of H₂, whose spectral line at 28 microns had never before been detected. The other was the detection of thermal water vapour emission, believed to be a prime coolant of cold shocked regions and, therefore, an essential factor in rapidly cooling shock-compressed gas on its way to protostellar collapse.

The S(0) transition was too faint to be previously observed even with the Kuiper Airborne Observatory, the only other far-infrared observatory equipped with powerful infrared spectrometers. The H₂ molecule is symmetric and therefore radiates only through quadrupole emission, which is orders of magnitude weaker than normal, dipole emission from asymmetric (polar) molecules like carbon monoxide, CO.

Because this H₂ line had never been detected, current estimates of the molecular hydrogen content of galaxies has been quite uncertain, though cold, molecular hydrogen is often a galaxy's dominant gaseous constituent. To date, estimates of the H₂ content of molecular clouds and galaxies has rested on determining the abundance of CO, and then inferring an H₂ content on the basis of presumed CO/H₂ ratios. Complicating the issue has been the C¹²O self-absorption, which meant that only the far lower, C¹³O abundances could be considered reliably established. But then, one required a double extrapolation to determine H₂ content, involving not only a guess at the CO/H₂ ratios but an additional uncertainty in the C¹³/C¹² fractions expected in molecular clouds.

To a significant extent these indirect determinations of molecular hydrogen abundance may now become a thing of the past, as molecular hydrogen detections were reported not only in Galactic sources, but also in several actively star-forming galaxies, such as M82 and NGC 253. In the ultraluminous galaxy NGC 6240, H₂ transitions as energetic as the S(7) line were detected in emission. In the galaxy NGC 6946, a first quantitative estimate of the H₂ content indicated as much as 80 Msolar pc^-2 of surface area, for the galaxy seen face-on.

These detections will not solve all our problems, since temperatures have to be as high as a few hundred degrees Kelvin before even the S(0) transition of H₂ can be excited, whereas temperatures deep inside molecular clouds generally are far lower. But at least it should now become possible to establish the ratio of H₂ to C¹²O or C¹³O in warmer regions of clouds and then extrapolate these ratios to colder domains to obtain the total mass of molecular hydrogen - the gaseous component responsible for star formation.

Equally exciting as these reports, were talks on the detection of water vapour emission, observed in any number of sources. The water vapour content of Earth's atmosphere is so great, even above aircraft and balloon altitudes, that the spectral regions in which cosmic H₂O strongly emits, are totally blocked. It took ISO, the first spacecraft to take infrared-sensing spectrometers totally beyond the atmosphere, to make observations of thermally-emitting H₂O. This radiation was detected in stellar winds streaming out from evolved oxygen-rich stars, in shocks emanating from Herbig-Haro objects, and in a variety of shock-compressed regions. Water vapour was also observed in absorption in dense clouds. Previously H₂O emission had been observed only in circumstellar and interstellar masers - rare, peculiar objects that offered no guide to the overall H₂O contents of typical Galactic clouds or external galaxies.

With the new water vapour observations in hand, some obtained with SWS, others with the Long Wavelength Spectrometer LWS, we should be able to answer a number of questions that have long remained open:

• Is rapid H₂O radiant-cooling a dominant factor keeping gas clouds compressed - preventing them from bouncing back - once they have been subjected to shocks. Shock-compression is believed to be one factor that can trigger protostellar collapse. Determining the role of H₂O in this process is, therefore, important.
• Equally important is the determination of the carbon-to-oxygen ratio in the interstellar medium. Previous studies have suggested that this ratio may be far higher in interstellar space than in the solar system or other regions where it has been unambiguously determined. A substantial fraction of interstellar oxygen was always assumed to be hidden in unobserved H₂O molecules. With ISO this component is at last coming under scrutiny, and the results to be gathered during the life of the mission will certainly advance our understanding of cosmochemical abundances and processes.

• In the Herbig-Haro source HH 54B, preliminary estimates suggested that CO emission from this shocked domain out-radiate H₂O, by a factor of 10:1. How that emission ratio varies from one shocked region to another, will tell us the extent to which CO and H₂O control shock cooling and potentially protostellar collapse.
• An answer to the abundance question appeared to have been brought closer through yet another reported breakthrough reported at the meeting, the detection of solid CO₂. Previously only seen weakly in an IRAS-LRS spectrum at 15 microns, CO₂ was detected by ISO in the interstellar medium, at a wavelength of 4.27 microns, previously also hidden by telluric absorption. With this and other detections, it was now becoming possible to see CO₂, H₂O, and CO, in both their gaseous and solid forms. Atomic carbon and oxygen abundances can also be directly observed through those atoms' fine-structure transitions. And solid carbon, in the form of graphite, has long been observed, with some confidence. With all these dominant atomic, molecular, and solid species in which carbon and oxygen occur thus directly observable - with O2 the only exception - we are a great deal closer to determining the true abundance ratios of oxygen and carbon in much of the interstellar medium, and thereby also the universe.
• A particularly valuable investigative tool that emerged as a spin-off from these considerations, came from a report on the gas-to-solid ratio for CO, CO₂, and H₂O observed with ISO. Solid grains composed of these three species evaporate at quite different characteristic temperatures. The relative abundances of the solid forms of these molecules in equilibrium with their vapours, therefore, provide a good estimate of grain temperatures in interstellar clouds. With ISO we are thereby gaining a thermometer for measuring the temperatures of grains in interstellar clouds

At this first ISO workshop there was far more, of course - enough to bring a gleam to the eye of each attendee. Among the many reports, totaling about eighty if one counts oral as well as poster sessions, were astonishing results of all kinds:

• There were beautiful pictures gathered in a 15 micron camera survey, which exhibited Galactic dust lanes so dense as to be opaque even at these very long wavelengths. Breath-takingly beautiful camera pictures were also shown of external galaxies, and dark, extended Galactic clouds. And detailed investigations of these dark clouds were revealing embedded sources, likely protostellar objects, in numbers appreciably higher than seen before. ISO's potential for elucidating star formation in these regions is exceptionally promising.
• Most interestingly also, the photometer was detecting regions from which radiation at 200 microns was the dominant emission. In particular, the Chameleon-1 region was roughly a hundred times brighter at 200 microns than at 60 microns, and cold dust, responsible for this long-wavelength emission was being found in many other sources as well, including the galaxies NGC 5266 and 5156. These two sources had been traversed in a telescope slew during which the photometer was conducting part of its 200 micron serendipity survey. Such observations are permitting us to come to a significantly better understanding than we had to date of the total energy budget of different types of Galactic and extragalactic sources. Their emissions can now be observed all the way from the gamma ray region, through the X-ray, EUV, UV, optical, IR, submillimeter and radio range, with this closure of one of the last remaining energy gaps around 200 microns.

This was an exciting workshop. ISO is fulfilling its promises to the astronomical community and then some!

Martin Harwit
12 June 1996

First ISO Results have been published in the A&A Special Issue, Volume 315, Number 2, pp L27-L400, 1996.