142 Proceedings of the Royal Physical Society. 



doubt the winds do in time bodily transfer to our shores 

 water which may have been at one time beneath the 

 Equator; but it is by no means clear that such a transfer 

 could take place in so short a time as a few months. The 

 same remarks apply also, though with much less force, to 

 the aerial components of the " Gulf Stream." These are, of 

 course, primarily set in motion by solar energy, and their 

 subsequent directions of movement are determined by very 

 complicated factors, one of which must be briefly noticed 

 here, on account of its bearing upon the formation of snow. 

 Aqueous vapour is lighter than atmospheric air, in the 

 proportion of 625 to 1000. Hence air charged with aqueous 

 vapour tends to rise, while dry air has a corresponding 

 tendency to fall. So at no great distance from an area where 

 a copious evaporation is in progress, there must always be an 

 area where there is a certain amount of down-drauoht of 

 the drier air from the upper and colder regions of the 

 atmosphere, — the two conditions are necessarily correlative 

 to each other. It is true that it is often difficult, in the 

 present state of our knowledge, to state exactly what would 

 happen under any given combination of circumstances like 

 these, but the broad principles undoubtedly hold good. 

 What we are concerned with at present, however, is the fact 

 that when aqueous vapour is condensed into rain it gives 

 out to the air around it exactly the same quantity of heat as 

 was used up in the first instance in converting the equivalent 

 weight of water into the gaseous condition. That is to say, 

 each pound weight of rain (a trifle under a pint) liberates 

 sufficient of its latent heat of evaporation to the air around it 

 as would suffice to melt five pounds of cast iron. And, as 

 the specific heat of dry air is '2669, a very large yolume of 

 air is warmed by the process. It follows from this that the 

 condensation of a single grain of aqueous vapour liberates 

 sufficient of the once-latent heat to warm one cubic foot of 

 air 7°*25 F. As a result of the condensation on the west 

 coast of Ireland, Dr Haughton estimated that the quantity 

 of latent heat set free by this means is equivalent to nearly 

 half of the heat derived from the Sun. A large part of this 

 latent heat is, of course, used up directly in raising the 



