2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I07 



increased mining and smelting operations in the surrounding region. 

 The station was closed in 1946. In table i, which follows, only Table 

 Mountain and Montezuma values are given, omitting Tyrone. 



The Montezuma values are weaker than usual in 1939 and through- 

 out 1940. Meteorological conditions at Table Mountain are inferior 

 to those at Montezuma, especially in the months March to June, in- 

 clusive, as shown by figure 7, Annals, volume 5, so that intervals of 

 weakness at Montezuma cannot be fully corrected by Table Mountain 

 results. 



With these explanations I now give in table i the lo-day and 

 monthly mean values of the solar constant of radiation from October 

 1939 to December 1945. The individual days' results were thoroughly 

 gone over by Messrs. Aldrich and Hoover and Mrs. Bond, of the 

 Observatory staff, and all the individual observations were scrutinized 

 with all the care that long experience suggests. All the statistical evi- 

 dences as to accuracy and the methods of checking and correction, 

 as described in volume 6 of the Annals, were employed, except that 

 the spectroscopic method of getting "improved preferred" values, as 

 described at pages 166 and 167 of volume 6 of the Annals was not 

 used. In table i are given the year and month in the first column ; 

 in the second and third columns the mean decade and monthly values 

 from Table Mountain (T) and Montezuma (M). For each month 

 the monthly means follow the three decade means. In the fourth 

 column are given the preferred mean values for decades and for 

 months. To save printing, only the last two figures are given, so 

 that all values are to be understood as prefixed by 1.9. For example, 

 for October 1939, 43 means 1.943 in calories per square centimeter 

 per minute, at mean solar distance, outside the earth's atmosphere. 

 I do not give here the number of days of observation for individual 

 decades at the two stations. However, in computing "preferred mean" 

 values these data, and also the grade of the observations at the two 

 stations, were considered. 



It was immediately apparent that though there is fair agreement 

 between prophecy and observation up to near the end of 1944, the 

 large depression of solar-constant values, prophesied from 1939 to 

 occur in 1945, did not occur. I thought it might be because the master 

 period of 273 months was incomplete in 1939. I therefore used the 

 additional values 1939 to 1945 with those preceding, as given in table 

 ^7, Annals, volume 6, to make an entirely new analysis, after the 

 manner described in pages 178 to 182 of Annals, volume 6. After 

 tabulating the values for each periodicity in several successive groups, 

 covering respectively successive intervals of time in order to test the 



