146 ANNUAL REPORT SMITHSONIAN INSTITUTION, 194 1 



and through the work of Struve, of the Yerkes Observatory, the in- 

 tensities of the calcium lines were being used as a measure of the 

 distances of the stars in which they were observed. The greater the 

 distance of the star, the greater the amount of interstellar gas through 

 which its light passes, and the stronger the lines. Several broad hazy 

 lines, apparently originating from interstellar gases but differing 

 greatly in appearance from the normal sharp lines, had also been 

 discovered by Merrill at Mount Wilson. 



Until recent months, accordingly, the situation was that the gases 

 of calcium, sodium, titanium, and potassium had been identified in 

 interstellar space but that the origin of several fairly conspicuous 

 lines still remained unknown. The identified lines all arise from the 

 atoms of the elements, and naturally astronomers searched for identi- 

 fications of the remaining lines in the atomic spectra of other ele- 

 ments. This led to no success, however. The possibility was then 

 considered whether the unidentified lines could arise from molecules 

 instead of atoms. As I have already said, under ordinary conditions, 

 molecules of compounds produce bands consisting of hundreds or 

 even thousands of closely packed lines as contrasted with the simpler 

 spectrum of relatively few lines arising from the atom. Under the 

 conditions of interstellar space, however, with extraordinarily low 

 densities and temperatures, the molecular spectrum might well be 

 simplified and even reduced to a few observable lines. The suggestion 

 that the broad diffuse lines observed by Merrill might have a molecu- 

 lar origin was put forward by several investigators, and in the 

 specific case of one of the sharp lines discovered at Mount Wilson a 

 tentative identification with a line of the common hydrocarbon gas 

 CH was offered by Swings and Rosenfeld. An identification resting 

 upon a single line, however, necessarily remained somewhat doubtful. 



The next step was taken by McKellar at the Dominion Astro- 

 physical Observatory. Applying to molecular spectra the principles 

 derived from a study of some of the identified atomic lines, he was 

 able to predict the positions of several additional lines for each 

 molecular spectrum. Thus if the single relatively prominent line 

 tentatively assigned to CH were correctly identified, there should be 

 at least three other fainter lines present in another region of the 

 spectrum. Similarly McKellar could predict the positions of certain 

 lines of the familiar cyanogen gas CN. 



Hence the final solution of the question came back to the observer. 

 Since the predicted lines were faint and narrow, it was clear that 

 photographic plates of high contrast and fine grain must be used and 

 that exposure times would be long. Fortunately a bright star was 

 available, Zeta Ophiuchi, lying near the southern Milky Way. The 

 100-inch telescope and a spectroscope 114 inches long were used with 



