"! 



., f 



G,38 



iu:Poin' — 1884 



inches, and in No. i?, O'.jDO inelics ; llie <jrra(luiition in buth Loiiii^- to tenths of a 

 depfi'ee. Th(\y were tested by iiunier.^inj^ them in a wooden vessel iilled with snow 

 and water, ])hiced in a room in which the temperature varied J'rom C4'' to .'57''. By 

 the means of several trials the corrections for the meltiii"- point of icc^ were founij 

 to be, for No. 1, ()'S7^ and for No. 2, 0'83''. No. 2 was found broken Feljvnarv 17 

 1S71, which brou^dit the experiments to an end. 'J'hirteen years afterwards No. 1 

 was test(id again in a similar maimer, and the correction was found to be 1 •02", tin- 

 change being O'lo°. One of these thermometers was placed in tlio liole in tlie 

 blocic of ice, and the oilier in the water in whicli it lloated, and to eliminati' errors 

 from imperfect corrections they were interchanged from time to time. 



The residl-s of these experiments are given in tlie following talile. '!'i varv 

 the conditions, the hole in tlie block of ice was tilled with diU'erent substimces as 

 stated in the talile. It was snp)iosed that tlie substance tliat would enable the 

 temperature of the ice to be ascertained witli the greatest precisien was benzine. 

 The mean of the two results in this condition when thermometer No. 1 wns in the 

 interior of the ice, and No. 2 in the water in which the ice floated, gavs^ the tem- 

 perature of the interior of the ice(>188° below that of the water in wliich it lloated ; 

 and the mean of the two trials, when No. 2 was in the ice and No. 1 in the watui', 

 gave the temperature of the interior of tlie ice, ()-2('i()' below that of tlie water in 

 which tlie ice lloated. Tlie mean of the four trials in which benzine w;is us;d 

 giving the temperature of the inleiior of the ice was 0'227 ' below thai of the water 

 in which it lloated. 



'J'he mean of the two results in wliicli air occupied the interioi- di" the ice, flio 

 thernuinieters being interchanged, gave the 1em})eralure of th'.' ice ()'07^^ below 

 that of the water in which it lloated. 



In the single result in wliich llie hole was filled with water, the readings gave 

 the temperature! of the interior 01)17 ' below that of the water in which the ice 

 floated. 



The mean result of all 1h(> trials gives the teui]iei'ature of the iiilinliu' of ihr 

 block of ice O'loir Ijelow that of the water in which it lloated; the irregiilai'llies 

 in the results of the several trials an; however too great to ])ennil (lie adoption of 

 this mean as the exact dillerence. i'lvery trial, however, gave a less diti'erence 

 than was found by Forbes, and I tliink it is safe to say that us a residt of the.-c 

 experiments, the temperature of t!i(> interior of a liledi of fresh water ice lloating 

 in iVt'sli water, the temperature of the air Ix'iiig between ;iO-r/' and lli'd" bahiviilieil, 

 is about one-fiftli of a degree below tliirty-two degrees, or the melting point of ice. 



During the above experiments, the block of ice became reduced by thawing from 

 its original diameter of seventeen inches to about fourteen inches. 



