Abstract
The author reviews the molecular opacity data that are available from various sources, and compares the methods that have been used in computing them. Line strengths, excitation energies and frequencies for 22 million lines of 20 different diatomic molecules and 35 million lines of 4 polyatomic molecules exist in astrophysical data bases, which is probably far from sufficient for interpretation of the accumulating amount of satellite data especially in the infrared. Also, data are lacking for good model atmospheric computation, and the quality and effect of a great fraction of the data on the weak lines have never been analysed. Practically nothing is known about the opacity of the bigger molecules that are observed or suspected in the circumstellar envelopes, and their possible presence in the upper photosphere may substantially alter our picture of cool stellar atmospheres. The author reviews the recent progress in the understanding of this new field and its possible importance for stellar atmospheres.
