54 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



From the observations on the nights of August 3, 4, 5, and 11 on 

 the summit of Mount Whitney (during these nights the observa- 

 tions were carried on continuously from evening to morning), I 

 have computed the means of the radiation, the temperature, and the 

 humidity, corresponding to different hours. The result is given by 

 figure 6, where the curve RR corresponds to the radiation; the 

 curves HH and TT to the humidity and the temperature, respectively. 

 The radiation decreases slowly from 9 o'clock in the evening to about 

 2 o'clock in the morning. At about 2:30 the radiation is subjected 

 to a rapid increase; between 3 and 4 o'clock it keeps a somewhat 

 higher value than during the rest of the night. The temperature, 

 which shows a very continuous decrease from evening to morning, 

 evidently cannot be regarded as a cause for these conditions. An 

 examination of the humidity conditions shows however that the abso- 

 lute humidity is subjected to a very marked decrease, which is per- 

 fectly simultaneous with the named increase in the effective radiation. 

 Considering that the previous investigations, discussed in this paper, 

 show that low humidity and high radiation correspond to one another, 

 we must conclude that the maximum of radiation occurring in the 

 morning before sunrise, is caused by a rapid decrease of the humidity 

 at that time. It seems very probable to me that the maximum obtained 

 by Exner from his observations on Sonnblick, may be explained in 

 the same way. 



E. INFLUENCE OF CLOUDS 



The influence of clouds upon the radiation processes within the 

 atmosphere is of very great importance for many meteorological 

 questions. At the same time the problem is an immensely difficult 

 one, because of the irregularities of the fundamental phenomenon 

 itself. Take the question of the influence of the conditions of the 

 atmosphere upon the amount of radiation reaching us from the sun. 

 When the sky is clear, we can probably calculate from a single obser- 

 vation, or a couple of observations, together with one or two known 

 facts, the whole access of radiation during the day to within perhaps 

 5 per cent. But as soon as clouds are present, we have to fall back 

 upon continuous observations, the occurrence and density of the 

 clouds, and the time of their appearance being subject to no known 

 general law that holds for such small intervals of time as we wish to 

 consider. Moreover the influence of clouds upon the solar radiation 

 is very great, the radiation being reduced to a very small fraction of 

 its former value by the interference of a cloud. Similar condi- 

 tions hold in regard to the effective radiation to the sky. As this 



