I50 CARNEGIE INSTITUTION 



If a new institution, provided with unusual instrumental means, 

 is to be established for the study of the sun, it should have the solu- 

 tion of one especially important and fundamental problem as its 

 principal purpose. The most important question seems to me to be 

 with regard to the constitution of the sun. We seem to be infinitely 

 far from the solution of this question — whether it is wholly gaseous 

 or in part liquid or solid. This question must be solved through 

 a ver>' detailed study of the various parts of the sun. It seems to 

 me particularly important to make a studj' of the sun spots. If 

 their spectra could be photographed in the greatest possible detail 

 throughout a solar C3xle with large gratings, an important advance 

 would probably be made. In this way it might be possible to form 

 a proper estimate of Julius's theorj' — either to confirm or refute it. 

 This investigation is a rather thankless task, since it consists only 

 in gathering observational material ; but a result drawn from it 

 would be far more important for our knowledge of the heavenly 

 bodies than the measurement of the radial velocity of a few hundred 

 stars or the intensity curves of variables. As soon as the charac- 

 teristic features of our sun are thoroughl}- understood, many other 

 phenomena will explain themselves. 



I would advise the use of plain gratings for this investigation, 

 since freedom from astigmatism is necessary ; the lenses should be 

 preferably of quartz-fluorspar, and the solar image on the slit should 

 be of large diameter. 



Terrestrial spectra must, of course, be employed for the explana- 

 tion of the phenomena. The laboratory' must, therefore, be sup- 

 plied with concave gratings of various numbers of lines to the inch 

 and various radii of curvature ; also with direct current for arc 

 lamps and alternating current for transformers, as well as induction 

 coils of various dimensions up to one meter spark length, in order 

 that the spectra of the elements ma}^ be studied under the most varied 

 conditions. 



I have touched upon only a single question, w^hich seems to me to 

 surpass all others in importance. It goes without sajdng that a 

 large institution would also undertake investigations requiring night 

 observations. I name here briefly only two of investigations which 

 seem to me particularly important : the spectra of the planets, in 

 order to determine the nature of their atmospheres, and the spectra 

 of a number of the brightest stars with a precision approaching as 

 closely as possible that of Rowland's solar spectrum, so that their 

 chemical constitution can be accurately' compared, and if differences 



