f 



^OF ARTS AND SCIENCES. 293 



pelled differs little, if any, from atmospheric air, is shown by the free 

 burning of the candle, as well as by the lack of any impression pro- 

 duced on the senses, except that arising from mere coldness. 



Considering that the drift heap in which the well is situated rests 

 evidently on limestone, and is not far distant from the junction of the 

 limestone witli the mica slate, or gneiss, we may easily conceive of 

 the occurrence of such caverns, fissures, natural conduits, and sub- 

 terranean water-courses as might complete an arrangement on tlie 

 principle of the water-trumpet, — one of tlie oklcst contrivances for 

 securing a blast to be used in iron-furnaces, — and thus afford a 

 constant and ever-renewed supply of condensed air. And, as the 

 ex[)eriments of Di\ Gorrie show that but a modei'ate degi-ee of con- 

 densation is necessary to enable air to become freezing cold by its 

 return to the normal bulk, we may be warranted in saying that such 

 a cause, though of moderate power and having various impediments 

 to overcome, would be sufficient to produce all the effects observed 

 in the case under consideration. Of course, the colder the air at first, 

 and the colder the compressing and cooling water, the greater would 

 be the refrigeration resulting from the subsequent expansion. And 

 the actual freezing must proceed with greater rapidity at that time 

 of the year when the accumulated heat of the soil is allowed the 

 freest radiation, together with the least chance of increase. In fact, 

 it is said to be a matter of yearly observation, that the well " begins 

 to feel the cold weather," and to freeze over in autumn long before 

 there are any heavy frosts above, and, indeed, while the ground is 

 still open for tillage. This, certainly, would seem to indicate a cause 

 continually operating with almost uniform force. 



Since our return from Brandon, we find that the senior editor of 

 the American Journal of Science,* in commenting on an account of 

 the frozen well at Owego,t makes the simple remark, that the phe- 

 nomenon might possibly be occasioned by the escape of compressed 

 air ; but how the air could be condensed, and how or why the effect 

 should follow from such a cause, he does not point out. It does not, 

 therefore, appear whether or not Prof. Silliman had in his mind a 

 theory similar to ours. Excepting this indefinite suggestion, we have 

 met Avith no hypothesis capable of explaining all the facts in the case. 



* Am. Journ. Sci» 1839, [1.] XXXVI. 185. 



t The Braiulon well was not in existence at that time. 



