REPORT OF THE SECRETARY 105 



On the whole, therefore, although the results are as yet far from 

 being entirely satisfying, these experimental forecasts of Mr. Clay- 

 ton's are promising enough to warrant further trial. New methods 

 are continually being devised and tried in making them. Mr. Roe- 

 bling has generously arranged to continue them until June 30, 1925. 

 As the work is purely experimental no detailed publication of it will 

 be made at present. 



Naturally, if the forecasts made by Mr. Clayton really represent 

 solar changes, he can not succeed unless good solar measurements 

 are supplied. As soon as we began to receive daily telegrams from 

 both stations occasional fairly wide disagreements of individual days 

 commanded attention. We felt it necessary, in studying the causes 

 of such disagreements, to revise again entirely the systems of little 

 corrections to solar-constant values which have to be made to allow 

 for the haziness and humidity of our atmosphere. This revision 

 could be made with more advantage because much additional data 

 had meanwhile accumulated. 



Mr. Fowle and Mrs. Bond have worked over this matter during 

 practically their entire time, which, however, owing to furlough, 

 was only about three months in Mrs. Bond's case. A new method of 

 determining these corrections has been devised by the director and 

 Mr. Fowle, which eliminates satisfactorily the influence of the solar 

 changes which have occurred. Hitherto this matter of solar change 

 superposed upon the small terrestrial sources of error which we de- 

 sire to eliminate has been very embarrassing. Of course, if one 

 could wait many years before proceeding to evaluate the terrestrial 

 effects, the solar changes, being independent or but loosely connected 

 with local terrestrial ones, would be eliminated in the mean of a mass 

 of observations. We can, indeed, after several years more of observ- 

 ing, finally proceed in this way. But wishing to make immediate 

 use of our results a new method of procedure has fortunately oc- 

 curred to us which permits us to avoid the interference of solar 

 changes altogether. The details will be published soon. 



As both to us and to the Chief of the Weather Bureau it seemed 

 unwise to publish preliminary values of the solar constant which 

 later on would have to be corrected, we have discontinued the fre- 

 quent publications of them in the Monthly Weather Review which 

 we have been accustomed to make for several years past. After we 

 come to a fully satisfactory' basis of systematic atmospheric coiTec- 

 tions, these publications may be resumed. 



Of the two solar-radiation stations, Montezuma, Chile, has proved 

 far more suitable to the purpose than Harqua Plala, Ariz. It seems 

 probable that a place somewhat farther west and decidedly higher 

 would be preferable to Mount Harqua Hala. Violent storms occur 

 there in various months of the year, and the summer months in par- 



