ON THE CONTRACTION AND EXPANSION OF ICE. 42) 
day and night, with great care, and every precaution taken which 
prudence could suggest to insure accuracy of result. 
Herewith is given a table of observations and readings of the 
graduated rod, from the 29th of January to the lst of March. 
In order the better to illustrate the ice movement, I constructed 
the accompanying diagram. The datum is time, the upper section 
shows the lineal contraction and expansion of a body of ice one hun- 
dred feet in length, as read from the graduated rod. The vertical 
scale is-eight and a half times that of the actual movement, the better , 
to exhibit the variations. The section immediately beneath the ice 
line represents the atmospheric changes, as indicated by the mercurial 
thermometer, (Fah.) to the same datum of time, and to a vertical 
scale corresponding to the latter ice movement. 
Were the ice equally as sensitive to changes of temperature, and as 
quick to move as mercury, these lines, if applied to each other, would. 
almost coincide. The lower line exhibits the thickness of the ice at 
different periods during the experiment. 
It will be observed, on referring to the upper section, that the ice 
exhibited no movement from the 27th January to 1 p.m. on the 29th ; 
although the temperature of the atmosphere varied considerably 
during this period; it was not until the ice attained a thickness of 
three inches that it became susceptible of atmospheric influences. 
The phenomenon may be explained, I presume, by supposing the 
temperature of the ice, while yet thin, to be controlled by that of the 
underlying water. 
The expansion and contraction of the ice from the 29th of January 
to the 9th of February is remarkably uniform, and exhibits its great 
sensitiveness to changes of temperature. 
The average movement per degree per foot during this period is 
-00000 330. This ice, forming under cover, and protected from the 
deteriorating mfluences of sun, wind, rain, and snow, and not having 
been subjected to a high or wasting temperature, until the 5th ult., 
may, I think, be correctly termed pure ice. 
The ice from noon on the 5th February until 10 a.m. on the 8th 
was, however, (with the exception of a short interval) subject to a 
temperature varying from 28° to 36° and and was consequently 
absorbing latent heat, which, of course, materially changed its cha- 
racter. The temperature on the eighth suddenly fell to zero, and the 
ice (as soon as its moist surface was consolidated) contracted at the 
