180 ANNUAL OF SCIENTIFIC DISCOVERY. 



Sometimes the ice expands several feet on the shore, without any fissures 

 being created in the lake : this is caused by a temperature of the atmosphere 

 higher than that which previously existed. If the thermometer indicates a 

 temperature of 30 below zero, and then suddenly rises to zero, expansion 

 of the ice results. When the thermometer indicates 30 above zero, and 

 then falls to zero, contraction of the ice is the result. The force with which 

 ice expands depends entirely on the extent of the change of temperature. 



The most forcible movements of ice occur previous to rain storms. A 

 sudden rise of 20 in temperature produces violent expansion. Strong oak 

 piles in the bridge, which would not bend, were cracked and splintered by 

 the ice expansion ; heavy cap timbers of pine were snapped like reeds, and 

 heavy iron rails were curved and doubled up, as if put into a huge press. 

 Trees growing on the shore have been torn up by the roots, by the ice 

 expansion, and boulders weighing several tons have been lifted from the 

 shore, and forced into the bridge timbers. On one occasion, the ice ex- 

 panded no less than six feet along the Avhole shore. A uniform temperature 

 of the atmosphere neither causes expansion nor contraction of ice; it mat- 

 ters not whether the temperature is high or low, no movement of any kind 

 takes place. A coating of snow six inches deep effectually prevents any 

 motion in the ice, as it is a most effectual nonconductor, and protects it 

 from the influence of the atmosphere. 



Ice does not possess the power of contraction to the same extent as that of 

 expansion. It has been noticed that when it expands some feet, it does not 

 recede when the temperature falls to its former situation; it only contracts 

 by inches for its expansion in feet. 



The following are the general inferences deduced by Mr. Dumble, from his 

 observations : 



1st. That ice is capable of expansion and contraction. 



2d. That ice (up to 32) expands with a temperature higher than that which 

 had just previously existed. 



3d. That ice contracts with a temperature lower than that which had just 

 previously existed. 



4th. That ice does not expand or contract with a uniform temperature. 



5th. That ice is susceptible of expansion to a much greater extent than of 

 contraction. 



Gili. That when ice is equally dense, thick, and glare, and equally acted 

 on by a heated atmosphere, it expands from the centre towards the circum- 

 ference. 



7th. That ice expands towards the line of least resistance. 



NEW FORM OF AIR-PUMP. 



The accompanying figure represents a curious air-pump proposed by A. 

 Gairaud, of France, to supersede the common piston air-pump. The agent 

 for producing a vacuum in this pump is mercury, acting by gravity; and 

 instead of a flap-valve, as in the air-pump, air-tight faucets are substituted. 



It is conceded by all philosophers that, with the common air-pump, the 

 rarefication of the air can be carried on only to a certain limit; the best air- 

 pump not being able to bring the column of mercury in the barometer 

 attached to it below one-sixteenth of an inch; and it is obvious that the air 

 in the receiver will not be able to raise the valve, on account of its rarity. 

 These defects are proposed to be removed by the mercurial air-pump which 



