148 CARNEGIE INSTITUTION 



tions. Supposing that you were to come to the conclusion that the 

 solar constant observed, even at high altitudes, varies with a sun 

 spot period, it would still be open to doubt whether the change is 

 not due to something that happens in the higher layers of the atmos- 

 phere. That, in niy opinion, would be the most probable explana- 

 tion, but as even this fact would be important to establish, I quite 

 agree that observations on the solar constant should be made at fre- 

 quent intervals. 



Whether there are any actual differences in solar radiations at 

 different times is a question that will, I think, be solved in a different 

 manner. It seems to me exceedingly unlikely that any increase or 

 diminution in solar radiation can take place equally and simultane- 

 ously all over the solar surface. If sun spots have anything to do 

 with it, we must imagine the changes to come out differently or at 

 least at different times in different solar latitudes. I would therefore 

 consider it a matter of first importance to improve observational 

 methods as much as possible, so as to be able to compare with the 

 utmost accuracy different parts of the solar surface in different por- 

 tions of the solar spectrum. A change in temperature of even ioo° 

 ought to make an appreciable difference in the radiation at the violet 

 end, though the radiation in the red will not be affected nearly as 

 much. I have been very much struck with a recent paper by Mr. 

 Wood, of Baltimore, which describes a screen that absorbs the visible 

 light, leaving the ultra-violet. I should say that simultaneous 

 photographs taken of the solar disc by the ordinary method and 

 with this screen might give interesting results. 



2. I think you know already the great importance I attach to the 

 careful investigation of the spectrum of sun spots, and the other 

 points you mention are of equal interest. 



3. Here again I agree with you that such work as the measure- 

 ment of the heat radiation of stars and investigations of spectra with 

 very high dispersion are bound to lead to important results, and if 

 your present atmospheric conditions at available observatories are 

 not sufficiently good it would no doubt be highly advisable to have 

 a site specially selected for the purpose. 



I am not, of course, able to judge whether the same site may be 

 suitable for the solar and stellar work, and I do not want to dis- 

 courage altogether the possibility of a station at the base of a moun- 

 tain on which solar observations are taken. Some useful results 

 might be obtained in this way, but I should not hope for very much 

 and should not personally be inclined to recommend any great ex- 



