4 Canon Moseley on the Mechanical Properties of Ice. 



the other bolt. The measurements being in each case taken, 

 and the heights of the thermometers recorded, the measuring-rod 

 was carried to the opposite vertical face of the block, and they 

 were there repeated. Corrections were thus applied for the dif- 

 ferences of expansion on opposite sides of the ice-block, and dis- 

 tortions in the form of the measuring-rod. Before the measuring- 

 rod was brought out, at each observation, it was compared with 

 the standard length on the wall, and when it was brought back 

 the comparison was repeated. 



The observations began on the 11th of February, and by the 

 continuance of frost M. Schumacher was enabled to go on with 

 them to the 24th of March, During this time he made 153 

 different admeasurements, each separate and complete in itself, 

 but all on the same ice-block. All are included in the mean of 

 his results. 



When he had completed 72 observations, M. Schumacher de- 

 termined to test the permanence of the temperature of the wall 

 in which were fixed the two bolts whose distance was his standard 

 of measure. For this purpose he bent the tube of a thermo- 

 meter at a right angle three inches above the bulb, and made a 

 corresponding hole in the wall, into which he plastered up the 

 bent part of the thermometer, but so that the bulb should stand 

 free of the masonry. The temperature of the wall was found to 

 vary as much as 3° R. 



It is not necessary to describe in detail the corrections made 

 by M. Schumacher for the resulting variation in his standard ; 

 or those by which he corrected other errors incidental to his 

 experiments, or the expedients he adopted in the construction of 

 his thermometers, or the corrections he applied to his readings 

 from them. They are models of ingenuity and laborious fidelity 

 in observation. As the final result he gives 



The coefficient of the linear expansion of) .nnnru^oo^ ux. t. t_i 



ice for 1° R. between the limits- 1°-6 R. I 000064236 > Wlth a P r °bable 



and-21°'8R | error of '000000193. 



The coefficient of the influence of the dif- . 000 002165, with a probable 

 ierence of the temperature of the air } £ .hnnnm i\ca 



e *x. 4. **m ui i iod ( error of '000001064. 



from that of the ice-block per 1 K. . . J 



The coefficient of the expansion of the I '000006636, with a probable 

 standard measure ( error of "000001027. 



The rate of Cooling of Ice. 



To determine the rate of cooling of ice, M. Schumacher brought 

 his block, when at a temperature of —2° R, and hung it in the 

 open air when at a temperature of — 17° R. He then observed 

 the descent of its temperature by the readings of the thermome- 

 ters frozen into it every five minutes. From these observations 

 it resulted that the sinking (At) of the temperature of the ice 



