324 ATMOSPHERE IN RELATION TO HUMAN LIFE AND HEALTH. 



writer lias been led by many personal observations to the conclusion 

 that the crystals are differently developed according to (1) the amount 

 of dust or nuclei in the air, (2) the electric state, (3) the humidity of 

 the stratum where they have their origin and of the lower strata, and 

 (3) the suddenness or slowness of their growth. He found that in a 

 clear air on a hill crystals on vegetation were clearer, simpler, and more 

 glassy than in the rather foggy neighboring valley; that in the neigh- 

 borhood (10 miles) of London, where the air was smoky, the crystals 

 on trees were very much more feathery, branching, and opaque, and 

 yielded smoky water on melting. The upper air varies greatly in the 

 amount of contained dust nuclei, in free electricity, and in differences 

 of temperature between strata. A moist southerly wind beating back 

 a cold northeast wind in England generally yields broad, heavy, irregu- 

 lar, conglomerated flakes; a dry gentle wind, with uniform conditions, 

 yields regular crystals, small and thin; a very dry and cold air in the 

 early days of a severe frost sometimes gives showers of pellets of vari- 

 ous sizes, roughly hexagonal or polygonal, very dense, thick, opaque, 

 and like a number of superposed plates. In March, and sometimes in 

 April, a soft hail or dense pellets of snow fall in showers with a north- 

 east or north wind, and dry air, the showers alternating with bright 

 sunshine. At great heights in the Alps, the snow in winter is small 

 and powdery; in summer the flakes are much larger. 



Hail is often the result of a sudden condensation of very warm, moist 

 air by great reduction of temperature at a great height. The dust 

 nuclei are soon all occupied by moisture condensed upon them, and as 

 the vapor falls to and below saturation point in a high column, it has 

 not sufficient nuclei on which to condense in cloudy form, and precipi- 

 tation takes place at a great rate, either on the cloud globules or on 

 the snow crystals which fall through from the upper part of the cloud. 

 Since the whole or a great portion of the column of the topmost cloud 

 is below the freezing point, the globules as they come in contact with 

 the falling crystals instantly freeze, and so the ctystal grows and falls 

 ever faster, accumulating bands of ice and snowy particles according as 

 the air is clear and saturated, or else densely cloudy, through which it 

 passes. The electric charge being much denser comparatively on a 

 large drop or crystal than on a small one, and the vapor pressure being 

 less, the hailstones grow very quickly, and since they fall rapidly 

 through very thick clouds, they add much ice by mere impact at their 

 base. The radial structure so often observed indicates the origin of 

 the hailstone from a radial snowflake or hexagonal plate. Hailstones 

 of large size are produced in circumstances of great electric disturbance. 



Sometimes a hailstone has been found with finely developed hexag- 

 onal ice crystals growing like stalactites from a matrix. Possibly the 

 attachment of a flat hexagonal crystal at a certain stage in the fall of 

 the hailstone and the action of electricity in the rapid passage through 

 the air are sufficient to account for these large ice crystals, but they 



