202 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 31 



The main key to the puzzle is that we can observe but a part of 

 the whole spectrum of the sun or of a star. Our atmosphere, fairly 

 transparent (in good weather!) for visible light and for the near 

 ultra-violet, becomes suddenly opacjue between 2,900 and 3,000 Ang- 

 stroms — that is, at about three-quarters of the wave length of the 

 ordinary limit of visibility in the violet. For all shorter waves it 

 is as opaque as a stone wall. This limitation, which cuts the astro- 

 physicist off permanently from the most interestin'g region of the 

 spectrum, arises from the presence of a small percentage of ozone 

 at the very top of the atmosphere, far too high to hope to penetrate 

 by any known means of flight. At the other end things are not so 

 bad. The eye ceases to be sensitive about A8,000. Modern photo- 

 graphic plates run out beyond AlO,000, with hope of future gain. 

 Energy-measuring devices, such as the bolometer used with such 

 success by Langley and Abbot, are free from this limitation and 

 have detected solar radiation beyond AlOO,000, but they are relatively 

 insensitive and record only a few of the strongest spectral lines. 

 Moreover, in the infra-red the water vapor and carbon dioxide of 

 the earth's atmosphere cut huge, wide bands out of the spectrum. 

 We may really count ourselves fortunate that the spectral region in 

 which it is easy to work is so little bedeviled by the influence of the 

 earth's atmosphere, which, as my old teacher. Professor Young, used 

 to say, translating literally from the French, " is the astronomer's 

 black beast!" 



At present our observations of the solar spectrum are impossible 

 below A2,900; accurate and complete from A3,000 to a7,300; under 

 completion to AiO,000, with hope of considerable extension, and 

 possible, but rough, far into the longer wave lengths. 



Now, there are some elements, such as boron, which have not a 

 single spectral line in the observable range, though there are plenty 

 in the farther ultra-violet and presumably some {not yet discovered) 

 in the infra-red. The same could have been said for phosphorus un- 

 til certain successful photographs Avere taken at the Bureau of 

 Standards a month ago (December, 1931). There is evidently no 

 hope of identifying such elements directly in the solar spectrum. 



For many other elements the strongest lines lie far out of reach in 

 the ultra-violet. This is true for the nonmetallic elements without 

 exception and for some metals — gold, mercury, cadmium, etc. The 

 limitation of our observing powers puts us at a great disadvantage 

 when it comes to such spectra. A large majority of the "absent" 

 elements belong here. 



There are several more elements of this group which have their 

 next strongest lines deep in the red, beyond the limit to which Kow- 

 land could photograph. Modern plates reveal these lines (in all cases 



