EXPLANATORY HYPOTHESES 483 



THE VOLCANIC HYPOTHESIS 



We turn next to the highly geological hypothesis of volcanic dust as 

 the cause of climatic variations. It owes its importance to Abbott and 

 Fowle/ working together on the one hand, and to Humphreys/ working 

 by himself on the other hand. These authors seem to have shown almost 

 beyond doubt that volcanic eruptions of the explosive type are respon- 

 sible for a slight lowering of the earth's temperature. Their results are 

 summed up in figure 1, which is taken unchanged from Humphreys. 

 The upper curve shows the amount of solar radiation which actually 

 reaches the earth, as measured by the pyrheliometer. The next curve is 

 that of sun-spots inverted. The third will be described in a moment, as 

 will also the fourth. At the bottom the chief explosive volcanic erup- 

 tions are indicated. In general the eruptions appear to be accompanied 

 or closely followed by a fall in the pyrheliometric curve, thus indicating 

 that the volcanic dust cuts off a small part of the solar radiation. The 

 fourth curve, marked '^temperature departures," shows the average 

 amount by which the temperature of a number of selected stations de- 

 parted from the normal from 1880 to 1912. The stations were widely 

 distributed, but were confined to equatorial regions or else to continental 

 interiors — that is, to the parts of the earth where the temperature de- 

 pends most directly on the sun and is relatively little influenced by winds 

 and currents. These, it will be remembered, are the regions where Arc- 

 towski finds his synchronous variations of temperature in a two and a 

 lialf year cycle. This cycle does not appear in figure 1 because this is 

 based on annual means, and the cycle is so short that it can. be detected 

 only by the use of monthly means. An inspection of figure 1 shows that 

 tlie temperature curve does not agree with. that of the pyrheliometer. In 

 1903 and 1912, however, both are at a minimum, which suggests that 

 they are infiuenced by a common cause. The low places in the pyrhelio- 

 metric curve in 1885 and 1891 occur one and two years, respectively, 

 after the minima of the temperature curve to which they are supposed to 

 give rise. Moreover, the maxima of the two curves do not agree at all 

 closely. The most probable conclusion seems to be that the presence of 

 volcanic dust in the upper atmosphere is of real importance in determin- 



it swings back and forth across its continent, or even out into tlie ocean, and carries 

 conditions of high or low temperature with it. Such pleions or anti-pleions may hxst 

 for years, and may diminish in force and again be rejuvenated. 



' C. G. Abbott and F. E. Fowle : Volcanoes and climate. Smithsonian Misc. Coll., 

 vol. 60, 1913, 24 pp. 



" W. .T. Humphreys : Volcanic dust and other factors in the production of climatic 

 changes and their possible relation to Ice ages. P.ulletin of the Mount Weather Obser- 

 vatory, vol. G, part i, 1913, 26 pp. 



