204 FKOZEN WELL. 



of frozen materials ? We confess that when a series of warmer seasons should succeed, 

 as they certainly would to balance the cold ones, we should expect the frost to disappear, 

 unless in some way the openings of the pervious strata should be more or less closed. So 

 that, while we suggest the possibility of such an origin to frozen buried masses, we cannot 

 feel that the hypothesis has so much of plausibility in it as to be very probable. We 

 have more confidence in the idea of the preservation of ice through the long interval since 

 the drift period, and that this, in the manner that has been pointed out, gives rise to the 

 yearly increase of the congelation, and retards the liquefaction in the summer. 



From the difficulty with which air is made to pass through obstructed passages even 

 greater than that of water we are aware that some doubt whether it passes at all 

 through gravel and sand. But since the fragments of such materials must be irregular, 

 and, therefore, have interstices between them, by what must the spaces be occupied if not 

 by air? If by air, it must be subject to the influence of gravity and of expansion and 

 contraction by heat and cold ; and, therefore, have some motion ; and if it can be made 

 to change its place at all, why not so as to form more or less of a current, through wide 

 spaces ? We regret that we have not found time to perform any experiments on the 

 transmission of air through pebbles and sand ; for we have been unable, with a single 

 exception, to find the record of any such experiments by others. The exception was an 

 experiment by Saussure, upon the effect of evaporation by air passing through a glass 

 tube only an inch in diameter, filled with fragments of wet stones. The air, when it en- 

 tered from a bellows, was at 72 ; but 66 when it emerged. 



A fact was mentioned in describing the Owego frozen well which suggests the existence 

 of subterranean currents of air. At the depth of thirty feet, not half the depth of the 

 well, the flame of a candle was deflected horizontally. It was not affected lower down, 

 and can, therefore, hardly have resulted from a descending current. Yet the effect upon 

 the candle was greater than we should expect from a current issuing from the side of the 

 well, though on a subject of which we know so little we can hardly tell what to expect. 



The not unfrequent occurrence of living frogs deep in compact gravel beds, proves the 

 existence, if not the motion, of air around them. For Dr. Buckland has shown that they 

 cannot live long in cavities hermetically sealed. 



An hypothesis to explain the frozen wells was advocated by Prof. Loomis, before the 

 American Scientific Association in 1860, which imputes the congelation to the descent of 

 cold air in the winter in consequence of its greater weight. The reason, according to him, 

 why in the vast majority of instances wells and caverns (for we understood him to explain 

 the lattter in the same way) do not freeze, is, that currents of water of rather high tem- 

 perature flow through them and give off heat. Now, in the first place, there is no 

 evidence that currents of water, with few exceptions, pass through wells, and not 

 commonly through caverns, unless with extreme slowness. In general the water below the 

 surface collects in the lowest places, and much more rarely than on the surface has any 

 current except to pour into a well when first dug, till the whole is brought to a level. 

 2. Ice caverns are nearly as rare as ice wells ; but by this hypothesis all of them not 

 traversed by streams- of water should be such. 3. Currents of cold air in summer in most 

 cases issue from, instead of descending into freezing caverns ; which shows that some other 

 cause produces the frost. 4. The congelation in these caverns is often greatest in the 



