368 HUMPHREYS: PRODUCTION OF CLIMATIC CHANGES 



it is easy to show that spherical particles of volcanic dust of the 

 size above determined would require more than a year to fall 

 from only the maximum height already reached by sounding 

 balloons down to the upper cloud limit. But as most volcanic 

 dust does not consist of solid spheres, but rather of flakes and 

 rods, and again, as much of it is finer than the size assumed, it 

 follows that the time of fall may, in rare cases, be as much as 

 two to four years, or possibly even longer. Obviously then vol- 

 canic dust once in the upper atmosphere must remain in it for 

 many months and be drifted out, from whatever origin, into a 

 thin veil covering perhaps the entire earth. Hence to find its 

 effect on the temperature of the lower atmosphere it is necessary 

 to determine its action on radiation, both terrestrial and solar. 



Comparative action of volcanic dust on terrestrial and on solar 

 radiations. Since those vqlcanic dust particles that remain long 

 suspended in the atmosphere are large in comparison to the cube 

 of the wave-length of solar radiation, at the region of maximum 

 intensity, and small in comparison to the cube of the wave- 

 length of terrestrial radiation, also at the region of maximum 

 intensity, it is easy, by the use of equations developed by Ray- 

 leigh,^ to compare the action of the dust on the two kinds of 

 radiation. 



This calculation shows that volcanic dust particles, of the size 

 indicated by Bishop's ring, is roughly 30 fold more effective in 

 shutting solar radiation out than it is in holding terrestrial radia- 

 tion in. Therefore a veil of volcanic dust must produce an 

 inverse green-house effect, and, if long continued, should per- 

 ceptibly lower our average temperature. Let us see then what 

 observational evidence we have on the effect of volcanic dust 

 on insolation intensity and average temperatures. 



Pyrheliometric records. This subject has been carefully studied 

 by Dr. Kimball^ of the U. S. Weather Bureau, who finds that 

 there was a marked decrease in the insolation intensity from 

 the latter part of 1883 (the year this kind of observation was 

 begun) to and including 1886, from 1888 to 1892, and during 

 1903. There has also been a similar decrease since about the 



Phil. Mag., 47:375. 1899. 

 * Bull. Mt. Weather Obsy., 3: 69. 1910. 



