REPORT OF COMMITTEE ON OBSERVATORIES 1 53 



\_From Professor C. E. Mcndcnhall, Physical Laboratory, University oj 



Wisconsin.] 



February 27, 1903. 



Your letter of February 2 came duly to hand, and I haye made a 

 few notes concerning that part of the proposed work with which I 

 am somewhat familiar. These I am very glad to send you now, 

 hoping that the delay has been of no inconvenience to you. I have 

 not thought that you wanted great detail, and have tried to avoid it. 

 Though most if not all of the suggestions are such as must have 

 already occurred to you, I nevertheless give them, thinking that 

 perhaps they may usefully serve to confirm if not to initiate. 



As regards the general scope of the work, I shall not presume to 

 speak further than to emphasize the importance of one point which 

 you mention, namely, provision for the study of such laboratory 

 problems as seem intimately connected with the solar and stellar 

 work. 



In deciding upon the site it may be well to keep in mind, besides 

 the primary requisites, the fact that the observatory will demand a 

 constant though small supply of power. Possibly in California some 

 long-distance transmission line or water power could be drawn upon. 



For the solar constant work examples of all the best forms of 

 pyrheliometers and actinometers should be provided and studied 

 with a view to improvement. No one of them seems independently 

 reliable at present, though the Angstrom compensation pyrhehometer 

 promises best. 



For detailed infra-red spectroscopic work there is no doubt that 

 the bolometer is the most immediately applicable, because of the 

 work of Abbot and others at the Smithsonian Astrophysical Observ- 

 atory. But the radiometer, thermopile, and radiomicrometer each 

 has its peculiar advantages, so that if they were properly modified 

 for linear spectroscopic work they might be used with advantage at 

 the two substations. It seems to me unlikely that they will ever 

 supplant the bolometer for accurate linear work in cases where the 

 best must be had at any cost. 



In designing the spectrobolographic outfit it seems to me ques- 

 tionable whether it should be as large as that of the Smithsonian 

 Institution. At any rate, even if one outfit of such size is provided 

 for the main station, it would be desirable to provide another of con- 

 siderably smaller size, more or less self contained and capable of being 

 used with advantage by a single observer. For some of the work 

 n 



