LOOMIS'S THEOEY. 205 



summer, the reverse of what it ought to be by this hypothesis. 5. But even if we admit 

 that congelation may be produced in some open excavations by the descent of cold air, yet 

 how could it produce such a thick mass of frozen earth as at Brandon, where from the 

 character of the formation there can be no cavern, and where layers of clay must prevent 

 the direct descent of air except in infinitesimal quantities. Though we believe that the 

 external air does work its way horizontally through the gravel beds in this place, yet, as 

 in the ice caverns, it is chiefly the evaporation thereby produced that causes the cold. 



These objections seem to us insuperable against this hypothesis,-nor would we with 

 a dogmatic spirit urge our own views, but throw them out rather as suggestions for 

 the consideration of scientific men. We have a suspicion that if careful and extended 

 observations were made upon the temperature of springs and wells, evidence would be 

 found in many places of the existence of a degree of cold in the earth explicable only by 

 some local cause, such for instance as the existence of remnants of the congelation of the 

 drift period. Is it not strange that this subject has received so little attention from geolo- 

 gists? Is it not deserving the attention of learned societies? we mean, that they should 

 institute observations upon the temperature of springs, wells and caverns, and upon their 

 geological situation; also upon the extent and character of subterranean air currents and 

 subterranean evaporation. Not unfrequently such observations might not only eclaircise 

 the phenomena of frozen wells and caverns, but afford important hints in the construction 

 of ice houses and as to other means of obtaining a low temperature in hot weather. 



We here insert the letter before alluded to, embracing the views of Mr. Hagar : 



THE FROZEN WELL OF BRANDON. 



Proctorsville, Vt., July 15, 1859. 

 PROF. HITCHCOCK : 



My Dea/r Sir, Agreeably to promise, I will give you my theory concerning the Frozen Well at 

 Brandon. A few facts concerning it, its position and the surrounding objects, will first be given, as they 

 form the basis upon which the theory is predicated. 



The frozen well, which was dug last fall, is situated on the land of Andrew Trowmby, about three- fourths 

 of a mile west of the village of Brandon, in a basin between two outcrops or spurs of limestone, that are 

 about one-fourth of a mile apart, and running nearly parallel with each other in a northerly and southerly 

 direction. Alternations of clay and gravel are common in the vicinity of the well. During the coldest 

 weather last winter, ice formed upon the water in it, to the thickness of two inches, in a single night. The 

 water, this summer, is not covered with ice, still, it indicates a temperature of only one degree above 

 the freezing point, and yesterday I distinctly saw an incrustation of ice, several inches in thickness, upon 

 the wall, just above the water. 



North of the well, about ten rods, a spring issues from the surface of the ground, which is, in this place, 

 as bigh as the top of the well. One hundred and twenty-five rods southwest, there is a well five feet deep, 

 in a valley. The frozen well is thirty-four feet three inches deep, and has in it twenty-two inches 

 of water, resting on a bed of clean pebbles. The spring and shallow well alluded to, have clay bot- 

 toms, and the water in them is not unusually cold. From information obtained of those who saw the well 

 before it was stoned up, it appears that it is in what you denominate "modified drift," and in appearance 

 resembled the banks of the gravel pit situated 130 paces northwest of it. As this gravel pit forms an im- 

 portant link in the chain of evidence favoring my theory, I will describe it quite minutely. It is by the 

 side of tile road leading from Brandon to Sudbury. For several years gravel has been taken from this 

 place by tho citizens of Brandon, for roads, &c. An excavation is made about ten rods long, and in some 



