170 Dr. H. F. Weber on the Specific Heat of 



perature, inasmuch as the exact amount of heat, W J, necessary to 

 raise the unit weight of water from 0° to T is unknown. Bosscha's 

 calculation from Regnault's results (carried out between 100° 

 and 200°) gives Wj = T + 0*0001 IT 2 ; while Jamin and 

 Amaury*, for temperatures between 0° and 75°, give 



WT=T + 0-00055 T 2 + 0*0000004T 3 . 



Until these anomalous results are cleared up by a new research, 

 water must not be used for the empirical graduating of the ice- 

 calorimeter. I have used calcspar as a standard. From the 

 results of careful experiments* made with pure substance by 

 means of Neumann's method, the specific heat of calcspar be- 

 tween 6° and 100° was found to be 0*2065, the specific heat of 

 water for the temperature 2° to 6° being taken as the unit. From 

 another seriesof experiments upon the changeability of the specific 

 heat of calcspar, 0*2061 was deduced as the average specific heat 

 of this substance between the points 0° and 100°. If one unit 

 weight (one gramme) of calcspar be cooled from 100° to 0° in 

 the ice-calorimeter, 20*61 heat-units are given up to the calori- 

 meter ; and the heat-unit is taken as equal to the amount of 

 heat necessary to raise through one degree one gramme of water 

 at a temperature between 2° and 6°. Having obtained this con- 

 stant, a piece of pure calcspar weighing 1*521 grm., heated to a 

 boiling heat T, was brought into the calorimeter, and the ensuing 

 displacement of the mercury thread measured. The movement 

 of the mereurv thread corresponding to one heat-unit was then 



N 

 No=: 1*521 0-2061 T ' The Valu6S ° f N ° f ° r five different 

 scale- tubes were as follows, 



Scale-tube . . A. B. C. D. E. 



millira. niillim. millim. millim. millim. 



N = 14*42 10*24 9*80 603 411 



referred to the mean bore of the tubes. 



By careful calibration the individual bores of the tubes (600- 

 900 millims. long) could he compared among one another ; by 

 means of the calibration Tables any actual displacement could be 

 calculated in terms of the displacement which would have taken 

 place had the scale-tube possessed throughout the average bore. 

 The position of the mercury thread was read off by means of a 

 millimetre-scale, so placed that the scale-tube covered the middle 

 of the millimetre-marks. In reading off the position of the 

 mercury thread the eye was placed so that the upper, under, and 

 middle parts of the mark, seen through the tube, formed straight 

 lines ; in this way parallax was got rid off, and the readings 

 * Comptes Rendus, vol. lxx. p. 661. 



