THE CAUSE OF RAIN. 95 



and saturate it with vapor, and then by cooling cause the vapor to 

 condense, the resultant water is deposited directly. If the receiver is 

 filled with air not cleared of its dust, the cooling of the mixture of 

 air and vapor provokes first the formation of a fog that marks the 

 presence of dust, because each particle of dust becomes a nucleus, a 

 center of condensation, for the vapor. Finally, if the cooling is 

 carried far enough, the water formed falls in very fine droplets, each 

 one of which incloses a dust particle. Mr. Aitken has succeeded in 

 counting these droplets, by introducing only a very small volume of 

 dusty air into the receiver and finally filling it with absolutely pure 

 air. He has thus found that the external air contains on the average 

 32,000 particles of dust per cubic centimetre after a rain of con- 

 siderable duration, and 130,000 particles in fine weather. There 

 are 1,860,000 particles in the same volume of air in the middle of a 

 room, and 5,420,000 particles near the ceiling. The figures look 

 fanciful, but they are exact, for they have been corroborated by 

 numerous consistent experiments and agree with the determinations 

 that have been made by other methods. 



As to the formation of rain, it should be observed that absolutely 

 pure air can not give either fog or drops of water when it is super- 

 saturated with vapor. If there were no dust in the atmosphere we 

 should have no clouds or rain. The sky would always be clear, and 

 the sun would shine uninterruptedly as long as it was above the 

 horizon. There would be no dawn or twilight, and day and night 

 would succeed one another instantly, without transition. Atmos- 

 pheric water would be deposited only when in contact with things, as 

 in Aitken's experiments, very much as dew is deposited. 



The causes of the formation of rain are evidently the same every- 

 where. The secondary conditions change only according to climates ; 

 but they vary so much that rains are distributed very unequally 

 over the earth. According to Desanis, the quantity of vapor con- 

 tained in a column of air as high as the atmosphere would give, in 

 France, a layer of water about four centimetres thick. Few rain 

 storms would furnish so much; but there are storms sometimes that 

 give much more. On August 17, 1888, seven centimetres of water 

 fell at Clermont in five hours; and September 12, 1875, the pluvi- 

 ometer measured ten centimetres for the whole day. Still more 

 copious rains fall in some tropical countries; at Purneah, in India, 

 eighty-nine centimetres have fallen in twenty-four hours. 



Mr. John Murray has calculated, from the charts of Elias Loomis, 

 that the quantity of rain falling every year over the whole earth 

 would form a bed of water averaging nine hundred and seventy mil- 

 limetres in depth. 



When we consider the annual quantities of rain in particular 



