1890.] the Validity of Persons Absolute Zero. 17 



the substance in degrees centigrade (the latter being identical with 

 that of the calorimeter) ; " Cal." represents the total calories evolved 

 or absorbed, and " Cal. per 1 gram " those evolved or absorbed per unit 

 weight of substance, a deduction having been made for that portion 

 attributable to the bottle and its thermometer, namely 2'223 X t t' ; 

 C or c is the heat capacity deduced, the range of temperature and 

 mean temperature to which it applies being given under T, measured 

 in degrees above or below the melting point of the substance iu 

 question. Such results as apply to ranges of temperature partially 

 embraced in other determinations are enclosed in square brackets, and 

 are used for deducing those values opposite to which no experimental 

 data appear. The two values given for the heat capacity at O r (the 

 freezing point of the substance) are those which would be deduced 

 from the determinations at the two higher and two lower temperatures 

 respectively. Where the determinations include the heat of fusion 

 the differences between the individual experiments (not the means) 

 are used in calculating the heat capacity ; those determinations are 

 divided into two series, A and B (see Table II), each of which is 

 differentiated separately (Table II continued), and the means of these 

 two series of values taken ; then the experiments A and B are taken 

 together alternately, and the two other series thus obtained give 

 another series of mean results, the mean of these two means being 

 tinally taken. The heat of fusion was determined from those experi- 

 ments in which the initial temperature was nearest to the temperature 

 of fusion. 



The general results are collected in Table A, where those in the 

 first five lines give the values for 1 gram of substance, and those in 

 the second five the values for a gram-molecular proportion of it. 



Details. 



Before discussing the general results, the following details may be 

 noticed : 



Sulphuric Acid. The value found for the liquid at 19'53 seems to 

 be rather too high, and this makes the value deduced for 26'74 too 

 high, and that for 41'25 too low. The value for has here been 

 deduced diagrammatically, the probable error (Table A) being deter- 

 mined from the errors of the duplicate determinations. Both the values 

 for the solid at the initial temperature of 16'8 appeared somewhat 

 anomalous, and they were consequently omitted in the calculations. 

 In taking the mean value for T = 0, a double weight has been assigned 

 to the value deduced from the determinations at the two higher tem- 

 peratures. The probable error (given in Table A) in this, and most 

 other cases, has been deduced in the ordinary way from those two 

 values. The heat capacities of neither the solid nor the liquid show 

 any signs of an abnormal increase as we approach the melting point. 



VOL. XLIX. C 



