REPORT OF COMMITTEE ON OBSERVATORIES 1 5 



First, we have the improvement of the telescope. Iu its applica- 

 tion to this purpose, we believe that this improvement has by no 

 means reached its limit. 



Second, we have the development of spectrum analysis — the study 

 of light radiations in the spectrum. Rapid improvement in the appa- 

 ratus which this method employs has been continuous for nearly 

 forty years, and seems to be more strongly marked now than ever 

 before. This is well illustrated in the recent development of the 

 spectograph and the spectroheliograph, and in the marked advance 

 brought about in the size and quality of diffraction gratings. 



Third, there has been developed within the last quarter of a cen- 

 tury the power of accurately investigating the distribution of heat 

 radiations in the spectrum of celestial objects, and especial^ in that 

 of the sun. An immense advance in knowledge has been recorded 

 in this way through the accurate representation of a large portion 

 of the solar spectrum containing rays invisible to the eye and beyond 

 the reach of the photograph. 



These facts alone might not be sufficient to fully warrant the 

 project of combining in one large establishment the harmonious 

 cooperation of these three methods of research developed to the 

 highest state of science at the present day, but they should certainly 

 be sufficient to excite interest in the question. 



The present state of our knowledge of the sun is surprisingly 

 deficient in comparison with that which would seem to be attainable 

 by the use of the most powerful appliances. The fact that these 

 appliances have not been brought to bear in anything like the extent 

 demanded by needs well understood affords no very hopeful outlook 

 in the immediate future that these needs will be met at existing in- 

 stitutions. What immense additions to knowledge are possible 

 through the intelligent invention and use of powerful instruments 

 of research in this line is well illustrated in the normal solar spectrum 

 produced by Rowland less than twenty years ago, by methods 

 remarkable for their simplicity, which at once superseded all the 

 laborious efforts in that direction resulting from the use of less 

 powerful apparatus during the preceding thirty years. We believe 

 that similar opportunities for progress in solar research are within 

 the reach of investigators of the present day through the provision 

 of apparatus of which the possibilities are now generally conceded. 



A distinct advantage might also result in the attempt to carry on 

 in one institution the three related methods of investigation — tele- 



