264 PETTERSSON, ON WATEH AND ICE. 



means of a test tube filled with warm water, which is dipped 

 in the fluid. If the quantity of mercury is large [about 7 or 

 <S kilogrammes], and the stirrer is vigorously moved, the index 

 of the thermometer ean be kept almost stationary by means' 

 of this artifice, ^ with exception of a few oscillations, which 

 only at very low temperatures amount to O^io C but ordina- 

 •rily do not exceed O^^.os C. On accoiint of the large capacity 

 of the dilatometer, the oscillations of the mercury index of its 

 scale-tube are not perceptible, the meniscus moves very regu- 

 larly över the iisual number of milliraetres (30 unto 50) for 

 every centigrade degree and then, two minutes after the thermo- 

 meter has indicated the state of eciuilibrium of temperature in 

 the mercury bath, attains a fixed position without any oscilla- 

 tions. - The excellent conductivity of air-free ice, which is 

 surrounded by a metallic rind, is indeed surprising. In the 

 vicinity of the melting point, however, the behavior of the ice 

 is totally changed. In the next chapter it will be proved, that 

 at these temperatures a rapid change of the volume of the ice- 

 takes place. Immediately below the melting point the abnormal 

 change of volume increases enormously, and simultaneously 

 the index of the dilatometer becomes extremely sensitive to 

 slight oscillations of temperature. Happily, it is an easy task 

 to maintain the temperature absolutely coustant for a long 

 time in this case, because we may dispense with the cooling 

 mixture, provided that the temperature of the air in the room 

 is kept a few degrees below zero. Then, by repeated introduc- 

 tion of a warm body in the mercury bath, we may compen- 

 sate the slight loss of heat, suftered by conduction or radia- 

 tion, and approach the real melting point by regular observa- 

 tions unto 0^02 or O^.oi C. 



The dilatation of the frozen water being ascertained by 

 one or two series of observations between — 18° and the mel- 

 ting point, it remains to determine the change of volume by 

 the melting of the ice and the expansion of the liquid water. 

 For this purpose b is made to dip in a cup containing a 

 weighed quantity of mercury, c is shut, b is opened and the 

 temperature of the mercury bath raised above the melting- 

 point of the ice. After several hours the instrument is allowed 



' The operation, like most of tliese experiments, requires the aid of an 

 assistant, Avho with a niagnifying glass reads off the scale-tube of the dila- 

 tometer, while the experimentor regulates the temperature and observes: 

 the thermometer with a kathetometer. 



^ The temperature was of course kept constant much longer. 



