FROZEN WELL. 201 



congelation produced so deep beneath the surface ? 2. By what means is the frost pre- 

 served from external and internal heat? 



In reply to these questions we have two suggestions to make. 



1. These frozen deposits may have been produced during the glacial period that accom- 

 panied the formation of drift, and continued far down into the subsequent epochs of modi- 

 fied drift. 



That the drift period, and certainly the first part of the modified drift period, were 

 characterized by intense cold, such for instance as now exists in Greenland, almost all 

 geologists admit. The deposit at Brandon is probably about of the age of what we call 

 moraine terraces, whose peculiarities we have supposed produced by stranded icebergs, 

 and that the gravel and sand among these were doubtless frozen : indeed in some cases 

 we think such ridges as that west of the well were formed by successive layers of ice and 

 gravel. True, the period since that time we are disposed to reckon by piling tens of 

 thousands of years upon one another. But if we can show how a frozen deposit, formed 

 at any past period, might be indefinitely preserved, it is no matter how long ago it was 

 first congealed. 



There are two sources of heat that would, without special conditions, melt a frozen 

 deposit situated like that at Brandon : one, the heat of the sun penetrating downward, 

 and the other, the internal heat of the earth permeating upwards. From the experi- 

 ments of Prof. Forbes at Edinburg, it appears that external or solar heat penetrates only 

 49 feet in trap tufa, 62 feet in incoherent sand, and 91 feet in compact sandstone. 



We should expect that pebbles and coarse sand would be poor conductors of heat, 

 because the interstices are filled with air. The above statement shows that trap tufa, where 

 the interstices are very small, is still poorer as a conductor ; and we incline to believe that 

 clay, especially the marly clay, such as that at Brandon, is as good to resist the passage 

 of heat as tufa. One of us (A. D. Hager) made a few experiments on this point with this 

 clay (which will be detailed in another place), that go to strengthen this opinion. The 

 frequent preservation of ice among large loose blocks during the summer as in the Ice 

 Mountains in Virginia and Vermont, above described might seem to show that the larger 

 the fragments, and consequently the interstices, the greater the non-conducting power. 

 But we do not believe that the non-conducting power of the fragments is the principal 

 means, as it is in a refrigerator, of freezing the ice ; but rather the evaporation, which 

 would be greatest among large fragments. 



All the excavations at Brandon show a mixture and irregular interstratification of marly 

 clay and pebbles. The bands of clay doubtless lie both above and beneath the frozen 

 mass, even though our section (Fig. 100) may not correctly represent the prolongation 

 of the clay and gravel seen in the road cut, already described. But in all the excava- 

 tions both gravel and clay occur : and how almost impervious to heat must such a coating, 

 20 feet thick, be ! It Avould not, however, completely protect the subjacent mass from 

 solar heat. But there is another agency still more powerful for this end, namely, evap- 

 oration, which we think has operated here, as we shall more fully describe further on ; 

 and we think that these two agencies, namely, non-conduction and evaporation, may have 

 preserved this frozen deposit for a very long period, from exterior influences. 



But how could the deposit have been protected against heat from beneath, which, 

 14 



