5^8 t)n the supposed Change 



I confess myself surprized that so intelligent a man 

 should not have observed the fallacy of this reasoning. 

 He takes the mean of seven or eight severe cold da3^s in 

 1782, which covered Charles River with ice, for the 

 standard by which to estimate the cold of 1630, and the 

 mean of the whole winter^ as the standard of cold in 

 modern days, by which to compare it. This mode- of 

 reasoning is all fallacious. In the first place, it is not 

 true that a mean degree of cold, answering to IS'' by 

 Farenheit, is necessary to keep Charles River covered 

 with ice. The effect would be produced with a much 

 less degree of cold. Let the mercury sink to 10° for 

 five days, and a bridge of ice would be formed. Then 

 l€t the cold relax, and the mercury rise to 30° for five 

 days. The mean temperature of the whole ten days 

 would be 20°, Yet in this case probably, the ice would 

 remain a solid bridge through the whole time, notwith- 

 standing the rapid tides in that river. And in fresh wa- 

 ter, where there is no current, the bridge would remain 

 a much longer time, and in much milder weather. In- 

 deed, I can prove that a river or pond of water may be 

 covered with twelve inches of solid ice, when the meati 

 temperature is not below freezing point. But I will not 

 rest the argument on calculations ; I appeal to facts. 



In November, 1797, commenced a series of severe 

 cold, altho the beginning of the month was as mild as 

 usual. Towards the close of the month, the Hudson, 

 Delaware and Connecticut were covered with solid ice ; 

 yet the mean temperature of the whole month, at the 

 exchange at New- York, was 38° 87 by Farenheit, al- 

 most seven degrees above the freezing point. This fact 

 exhibits the fallacy of the Doctor's conclusions.... iSe^^. 

 59, of his History of Ferment. 



In page 383, Appendix, Dr. Williams states that in 

 America, where the rivers are froze to a firmness suffi- 

 cient to sustain heavy loads, the " mean heat of the win- 

 ters is from 15 to 20 degrees." This is a most egre- 

 gious mistake, and contradicts his own observations of 

 the weather between 1780 and 1788, as before stated. 

 The mean temperature of those seven winters was, by 

 his own statement, 25? 2 — and this corresponds nearly 



