xii 

moisture on it from the vapour would be large, there is 
no difficulty in explaining the size of all ordinary hail- 
stones. 
It must not be considered, however, that an ascending 
current is steady. Just as we have gusts and lulls in 
horizontal winds, so we have increases and decreases in 
the velocity of ascending currents. Thus a hailstone 
which has penetrated into the lower part of the cloud 
might be blown upwards and so go through the whole 
process again. In this case we should have a layer of 
white ice deposited around the clear layer, around 
which again there would be another layer of clear ice. 
In fact, if a hailstone is held by the ascending currents 
near to the region where the temperature is 0° C., it 
might well be carried up and down between the regions 
where the water is super-cooled and where it is not 
several times. We should then have several con- 
centric layers of clear and white ice, and a broken stone 
would have the appearance of an onion. Such cases 
are not at all uncommon, 
For the formation of hailstones two conditions must 
be fulfilled. 
(a) The clouds must extend through a great vertical 
height so that the three regions of water par- 
ticles, super-cooled particles, and snow are ex- 
tensive and well developed. : 
(b) There must be violent ascending currents, other- 
wise the stones would fall too rapidly to grow 
to a large size. 
These conditions are best fulfilled in warm regions, for 
there violent ascending currents are most easily de- 
veloped, and the condensation starts at a relatively 
high temperature, so giving regions of water particles 
and super-cooled water particles of great depth. 
These are the reasons why hailstones only occur during 
the summer in temperate regions, and why the most 
violent hailstorms and the largest hailstones are found 
in tropical regions. 
Soft Hail—If{ the temperature at the ground were 
much lower than in the case just considered, the region 
between the bottom of the cloud layer and the zero 
isothermal would be much reduced. It is in this 
region that the hailstone receives its coat of clear trans- 
parent ice. The hailstones which then fall would be 
relatively small, and consist only of the soft white 
balls appearing in the centre of the more complete 
hailstones. 
Falls of soft hail of this nature are quite common in 
the winter in Europe and in the hills of India; there 
are frequently falls of soft hail during the winter and 
spring in Simla. The reason is clear, for in Europe the 
temperature of the ascending current is so low that the 
freezing point is reached almost at the bottom of the 
Supplement to “ Nature,” April 14, 1923 
cloud, while in Simla the clouds form over the plains, 
and Simla itself is so high that the region in which water 
particles exist is mainly below the station. Thus the - 
hail which falls in Europe during the winter and in the 
hills of India falls almost immediately out of the region 
of super-cooled water particles, and therefore has had no 
opportunity for building up a layer of transparent ice. 
The form of these soft hailstones is most instructive. _ 
In most cases they are like cones with a hemispherical 
base. It is clear from this form that as they have 
fallen through the super-cooled region they have struck 
the water particles on their under sides. This has 
caused the bases to grow, and the cone above is really 
the shape of the stream-lines behind the enlarged base. 
A stone which has once commenced to have this form 
will retain it throughout, for the cone acts as a kind of 
tail and tends to keep the base always at right angles 
to the relative air motion. 
Snow.—Snow which forms over an ascending air 
current in which water particles first form will prob- 
ably have’ solidified cloud particles for nuclei. What- 
ever the auclei may be, as soon as the initial crystals 
are formed further condensation takes place exactly 
as in the precipitation of water, but the vapour 
condenses directly into the solid state without first 
going through the liquid state. The crystals of water 
are hexagonal prisms, and water in the crystalline state 
in the atmosphere shows all the wonderful shapes that 
this form of crystallisation can take. Having once 
started, the crystals may grow either along their central 
axis, giving rise to long thin prisms, or along their six 
axes to form hexagonal plates. 
Sometimes the growth is uniform, so that the result 
is a perfect hexagonal plate, at others the growth along 
the axes is more rapid than in the space between; this 
gives rise to stars, having a beautiful feathery appear- 
ance. The actual crystals vary in size, from minute 
crystals which can scarcely be seen by the naked eye to 
plates a quarter of an inch in diameter. In cold regions 
the crystals are small, because there is little water 
vapour present from which they can grow. In the 
Antarctic during the winter, when the temperature was 
always near or below o° F., only the smallest crystals 
were seen, so small that they were almost like dust. 
When crystals form at temperatures near the freezing 
point they grow to their largest size. When the air 
is full of large crystals frequent collisions take place. 
The crystals become interlocked and bundles of many 
separate crystals are formed ; these produce the ordi- 
nary snowflakes which, on account of their size and 
weight, fall relatively rapidly. It is to these latter that 
the term snow should be applied. With this restriction, 
snow occurs only when the temperature is near the 
freezing point. 
