184 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 70 



for example, that the variations in the temperature of the surface 

 and of the air in the middle of the Indian Ocean (see figs. 55, X-XI) 

 agree with the variations of the air temperature at the tropical 

 Stations of Mauritius, Batavia, and others (see figs. 68, 71). This 

 shows that the explanation of the general phenomena of tempera- 

 ture variations of the earth which we have referred to is untenable. 

 We shall later return to this consideration. 



With regard to the theory of Abbot and Fowle and of Humphreys, 

 that the extension of volcanic dust in the atmosphere has an im- 

 portant influence on variations of the temperature of the earth's 

 surface, the curves given by these authors of the pyrheliometric 

 measurements of the heat obtained from the solar radiation at the 

 surface of the earth do not fully prove their hypothesis, since the 

 curves have only a small similarity with the curve of the variations 

 in the yearly temperature of the earth. This latter has indeed a 

 very great similarity with the curve of sun spots. However, it 

 must be admitted that these authors have made it probable that 

 the volcanic dust which is distributed in the atmosphere, particu- 

 larly at times of the most violent volcanic eruptions, acts in such 

 a way that the temperature at the earth's surface is depressed, and 

 according to Humphreys' opinion it may be even possible that this 

 effect was in former times very considerable. But the influence 

 is not sufficient in order to explain the continuous and often great 

 variations in the climatic temperature of the earth. 



Humphreys' theory that the eleven-year variation in the tem- 

 perature of the earth, which is associated with sun spots, depends 

 on variations in the ozone formation in the atmosphere, assumes 

 a corresponding variation in the relation between incoming radia- 

 tion and outgoing radiation at the earth's surface, in other words, 

 the corresponding variation both in the daily and the yearly tem- 

 perature amplitude of the earth. But as we shall see later, such a 

 variation in this amplitude cannot be proved with certainty, at 

 least not such as the theory assumes. 



Finally we come to those theories according to which the variations 

 in the solar activity produce primarily variations in the air pres- 

 sure and in the circulation of the atmosphere which in their turn 

 influence other meterological elements. This view, which has been 

 advocated particularly by the two Lockyers and by Bigelow would 

 appear reasonable, but hitherto has had comparatively little support. 

 It agrees in its principal features with the results to which we have 

 arrived, and we shall refer later to this theory. 



