NO. 12 MOUNT ST. KATHERINE STATION ABBOT 9 



as having occurred in the year 1924.° Many short interval fluctuations 

 are also supported by both stations. 



There is little to choose between the two stations as to the smooth- 

 ness (i. e., freedom from wild values) of their curves during this 

 interval of 5 months. St. Katherine yields a few more days than 

 Montezuma. As the average daily discrepancy between the stations 

 during these 5 months is only 0.007 calorie, we infer that the average 

 accidental error of a single station then was but 0.005 calorie, or 

 5 of I percent. 



A careful record of conditions at St. Katherine has been kept 

 under Mr. Zodtner's direction. It is set forth in table 4. 



On the whole it appears that Mount St. Katherine, except for a 

 greater average wind velocity during the usual hours of observing, 

 and greater haziness during the spring months, is equally as favorable 

 a station for solar-constant observations as Montezuma. Moreover, 

 the two drawbacks just mentioned do not seem to have lowered the 

 quality of the daily St. Katherine observations below those of Monte- 

 zuma. Thus far her numbers of days of good observing quality per 

 year have slightly exceeded those of Montezuma, so that St. Katherine 

 may be ranked quite as high on the whole as Montezuma. 



In my paper " Weather Dominated by Solar Changes ", cited above, 

 I indicated the dependence of weather on solar fluctuations of short 

 interval. The results seemed to point to a possibility that at least at 

 some stations and during some months of the year forecasts of weather 

 for 10 days in advance might profitably be based on solar-radiation 

 observations if these could be of sufficient accuracy and continuity. 

 At that time it seemed doubtful if stations could be found whose 

 combined results would yield on nearly every day in the year solar- 

 constant values accurate to i of i percent as regards accidental error, 

 which seemed to be the minimum requirement for the purpose sug- 

 gested. But not only do our two best stations now nearly reach that 

 desired accuracy, but the impending substitution of Angstrom type 

 pyrheliometers at both stations as secondary instruments will prob- 

 ably increase this accuracy. 



For nearly 20 years H. H. Clayton has worked assiduously on the 

 problem of the correlation of solar variation with weather. In a 

 recent paper ' crammed with statistical results, he states : " In short, 

 these extensive data, covering all parts of the world, prove that solar 



- See Smithsonian Misc. Coll., vol. 85, no. i, fig. 3, 1931. 

 ^ Clayton, H. H., World weather and solar activity. Smithsonian Misc. Coll. 

 ,'ol. 89, no. 15, p. 24, 1934. 



