The ‘American Journal of Science’ (xxxvi. 184) gives a curious instance of a freezing-well near the village of Owego, three-quarters of a mile from the Susquehanna river.  The depth of the well is 77 feet, and for four or five months in the year the surface of the water is frozen so hard as to render the well useless.  Large masses of ice have been found in it late in July.  A thermometer, which stood at 68 deg. in the sun, fell to 30 deg. in fifteen minutes at the bottom of the well; and the men who made the well were forced to put on thick clothing in June, and even so could not work for more than two hours at a time.  No other well in that neighbourhood presents the same phenomenon.  A lighted candle was let down, and the flame became agitated and thrown in one direction at a depth of 30 feet, but was quite still at the bottom; where, however, it soon died out.  The water is hard or limestone water.

Rocks of volcanic formation would seem to afford favourable opportunities for the formation of ice.  Scrope mentions this fact in an account of the curious district called Eiffel or Eifel, in Rhenish Prussia, which was published originally in the ’Edinburgh Journal of Science,’[147] and has since been translated in Keferstein’s Deutschland.[148] The village of Roth, near Andernach, is built on a current of basalt, derived from the cone above it, which has at some time sent down a stream of lava to the north and west.  A small cavern near the village, forming the mouth of a deep fissure in the lava-stream, half-way up the cone, displays a phenomenon which the writer says he has often observed in volcanic formations.  The floor of the cavern was covered with a crust of ice at the time of his visit, about noon on a very hot day in August.  The peasants report that there is always ice in summer, and never in winter, when the sheep retreat to the cave on account of its warmth.  Steininger[149] found a thickness of 3 feet of ice on September 19, 1818, but it was evidently in a melting state, and the thermometer stood at 36.5 F. in the cavern.  He describes it as possessing a narrow entrance facing north, entirely sheltered from the sun by lava-rocks, and by the trees of a wood which covers the cone of scoria.

Scrope believes that this is the mouth of one of the arched galleries so frequently met with under lava in Iceland, Bourbon, and elsewhere; and on this he founds his explanation of the phenomenon.  If the other extremity is connected with the external air at a much lower level, a current of air must be constantly driven up this gallery, and in its passage will be dried by the absorbent nature of the rock—­which is perhaps partly owing to the sulphuric or muriatic acid it contains[150]—­and the evaporation caused by this current produces a coating of ice on the floor of the grotto, where there is a superficial rill of water.  The more rarified the lower external air, the more rapid will be the current of cool air; and, therefore, the greater the evaporation.  The winter phenomenon is to be explained by the fact that the current of air will be about the mean annual temperature of the district, taking its temperature, in fact, from the rocks through which it passes; and, therefore, by contrast the grotto will appear warm.