680 
mained in this region till 9.30 p.m., when the sun set. He 
was quickly enveloped in a thick mass of thunder clouds : the 
lightnings flashed on all sides, and the loud claps were 
incessant. The Thermometer again stood at 21°, and 
snow and sleet fell copiously. The storm lasted three 
hours. M. Testu says his balloon was affected by a sort of 
undulatory motion upwards and downwards, owing, he thought, 
to the electric action of the clouds. The lightning was 
excessively vivid ; the thunder, sharp and loud, preceded by 
a sort of crackling noise. In this account we have a 
uniform initial temperature for the electric rain of 66° 
Fahrenheit, and a uniform temperature of 21° Fahren- 
heit in the electric region through which the drops 
descend to earth, and when these considerations are 
joined to the rapidity of their motion, evaporation alone is 
abundantly sufficient to account for their conversion into 
ice. 
Were this an adequate explanation of what takes place, 
all the hailstones in a given storm should be alike, whereas 
they probably never are so, and in fact were not in the storm 
I have described. Other considerations therefore are 
required to account for the phenomena. What has been said 
would adequately explain hailstones of class I, or of class 
III, falling alone, but does not appear sufficient to account 
for their falling together, still less for those of class II, even 
if these only had fallen during the storm, and when the three 
descriptions fall together, the explanation is manifestly more 
insufficient. We must reconsider the effects rightly to appre- 
hend their causes. Solid hailstones are obviously drops of 
water formed by an aggregation of atmospheric vapour, 
which has been subjected to an evaporation sufficient to 
abstract the heat of fluidity and allow of solidification. The 
same effect would result from a moderate evaporation, long 
continued, as in the case of freezing water by dried oatmeal 
