190 ANNUAL OF SCIENTIFIC DISCOVERY. 



difference of expansion between mercury and glass. A boiling point and a 

 freezing point being duly established, the space between them should be 

 divided into equal parts, and the points between them constitute permanent 

 landmarks for the indication, not the measure, of temperature. 



Two points are still open to discussion : the size of the degrees, and the 

 zero. The size of the degrees should not be larger than the least now in use. 

 It would be better for the purpose of graduation that the number between 

 boiling and freezing were 256, or some oOter power of 2; but the mechanical 

 difficulty and the necessity for exactness are not so great as to demand a 

 change from 180, which makes them so small that we never divide them, ex- 

 cept in mean temperatures. Those of the Centigrade, and still more those of 

 Reaumer, are seriously objectionable, as observations ought generally to be 

 recorded in fractions. As to the zero, it should not be higher than the lowest 

 in use, as even that requires negative signs in observations on the air in high 

 latitudes. It is a pity that a zero were not taken near the freezing point of 

 mercury, or even below the supposed temperature of the planetary space ; but 

 the number of observers in high latitudes is too few to render a change ad- 

 visable for their convenience. But a zero at the freezing point is a serious 

 evil, involving so much labor in placing signs of positive and negative, and so 

 much error in the use of them as justly to demand a new scale were there 

 none better. 



But has the Centigrade scale no advantage over that of Fahrenheit in virtue 

 of its centesimal division? I answer, none whatever absolutely none. 

 Decimal sub-divisions are a great convenience, and it is to be hoped that the 

 first examples of each the American coin and the French weights and 

 measures may soon become universal but such convenience is not promoted 

 by the Centigrade scale, which has no subdivisions. For every day thousands 

 multiply and divide by 12, because 12 makes a dozen, 12 lines make an inch, 

 12 ounces a pound, and 12 pence a shilling; no one ever yet multiplied by 32, 

 because it is the freezing point, or divided by 212, because that is the boiling 

 point. Let us then adhere to our present scale so long as it shall be the best 

 extant, and not, by grasping at the shadowy advantages of a centesimal di- 

 vision, lose the substantial good of small degrees and a zero below the ordinary 

 range of atmospheric temperature. 



ON THE FROZEN WELLS OF. OWEGO, N. Y. 



At the last meeting of the American Association, Professor Brocklesby read 

 accounts of frozen wells in Tioga County, New York, one of which, situated 

 in a table-land, 30 feet above the level of the Susquehanna, has ice in it the 

 year round. The well has been dug 21 years ; depth from surface to the ice 

 61 feet. Ice could not be broken by a heavy weight attached to a rope ; 

 thermometer descended 38 in 16 minutes; men can not endure to work 

 more than two hours in the well, and a candle let down shows a deflection of 

 the flame in one direction, proving the existence of a current of air. A large 

 piece of ice was drawn up July 25, 1837. No other wells are within 60 or 

 80 rods, and none present similar phenomena. A gentleman in Owego 



