2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 7/ 



most favorable localities were chosen for the observatories. Our own 

 country does not afford the best localities in the world for this pur- 

 pose, for clouds and other variable atmospheric conditions are objec- 

 tionably prevalent even in our great southwestern deserts. Since 

 October, 1920, Mount Harqua Hala, in xA.rizona, has been continu- 

 ously occupied by us for the solar radiation work which before that 

 date was carried on during part of each year at Mount Wilson, 

 California. 



Fortunately, the Institution was able to devote to the work the 

 income of a portion of the Hodgkins fund, and therewith established 

 in 1918 another station near Calama, in the nitrate desert of Chile, in 

 South America. Our experience recommends this region very highly 

 for the purpose. But the combined support of Congressional appro- 

 priations and the Hodgkins income would have been insufficient to 

 maintain these two stations in the United States and Chile had not 

 Mr. John A. Roebling, of Bernardsville, N. J., helped in the work 

 very generously. By his advice and assistance, the Calama station 

 was removed in August, 1920. to Mount IMontezuma, nearly 10.000 

 feet high and about 12 miles from Calama. 



Mr. Roebling has not only provided many expensive comforts and 

 necessities for both stations, borne the heavy costs of transporting 

 observers, and supplemented their meager compensations to enable 

 them to bear the sacrifices involved in years of isolation, but he has 

 also engaged Mr. H. H. Clayton (with assistants) for the past two 

 years in the study of the effects of solar changes on weather and 

 climate. To promote this investigation, Mr. Roebling has provided 

 means for daily telegraphic advices from each observatory to the 

 Smithsonian Institution, and for telegraphing the mean values to 

 Mr. Clayton for his studies. 



In this way experimental forecasts have been prepared by Air. 

 Cla}'ton and communicated in advance to the Smithsonian Institution 

 so as to make possible an unbiased test of the value, if any, of knowl- 

 edge of solar changes for weather forecasting. As the results of these 

 experiments are soon to be published by Mr. Clayton, they need not 

 be mentioned here, further than to say that thev are of high interest 

 and promise. 



The daily telegrams from both stations, required for Clayton's fore- 

 casts, called attention so sharply to all discrepancies, that we found it 

 necessary to make several extensive investigations of sources of error, 

 methods of reduction, etc. These investigations are not yet finished. 

 We believe, however, that future modifications which mav result from 



