854 Kohlrausch and Leomis—Influence of Temperature 
atures for the several periods of vibration, i.e. they give the 
reading of the thermometers for the mean moment between the 
beginning and end elongation. When the interval of time is so 
considerable that the corresponding curve cannot be regarded as 
a straight line, the temperature t would be thus obtained some- 
what too low. To determine the true value, the arithmetical 
mean t, was taken from the temperatures which correspond to 
the beginning and end time of elongation. The correct tempe- 
rature is then 7+" Amcgh 
Correction of the readings of the thermometers.—The numbers 
tached to each Pog enacd of the wire extend so far within the 
ifference of temperature which might 
_ The above two conditions, however, must be fulfilled, or 
gi they are not fulfilled, the readings must be corrected. 
e space so completely as can be assumed of the fine wire. 
Now according to well known laws, the amount of heat lost by 
indicated by the thermometers, to be able to mdditlake the dif- 
nce, 1. e. the correction of the readings. For this determl- 
nation a very delicate thermometer was employed, whose cylin- 
drical quicksilver reservoir possessed a diameter of only some 
3™™, and contained in all less than 1% gr. of quicksilver. It 
