16 Mr. E. H. Griffiths and Miss Marshall on the 



galvanometer showed that all evaporation had ceased, the 

 dropper was extracted by means of a bent wire, immediately 

 placed in its case, and weighed again after standing some hours 

 in the balance-case. 



Let m 1 and m 2 be the weights of the case and dropper 

 before and after an experiment ; the temperature of the 

 balance-case was usually about 20° C, the specific gravity 

 of benzene at that temperature may be taken as 0*88, hence 

 M (true mass corrected to vacuo) 



= m x — m 2 + (m 1 — m 2 ) X "0012. 



11. Before commencing the experiments it was necessary 

 to ascertain if any alteration had taken place in the values 

 of the various constants and variables since their determination 

 in 1894. 



Thermometers II. and A were re-standardized and it was 

 found that a " zero-point rise " had taken place, as is customary 

 with mercury thermometers. 



Thermometer No. II. had risen 04 mm.=0 c, 02 G. since 

 its standardization in August 1894. 



Thermometer A had risen 1*4 mm. = o, 06 (nearly) since 

 its standardization in July 1893, which is about the normal 

 rise of thermometers of this description. 



The platinum thermometer (AB) was removed from the 

 calorimeter, strapped to its corresponding thermometer (CD) , 

 and placed in the tank whose temperature was raised from 

 18° C. to 40° C. Observations at different temperatures 

 showed a rise of 0*45 mm. in the null-point whose position is 

 now given by the formula 598 , 8 + , 03# in place of 598-35 + 

 •03 6. If, however, this change, which probably took place 

 in the arms of the bridge rather than in the thermometers *, 

 had not been detected the resulting error would have been 

 negligible, for a difference of *45 mm. in the setting of the 

 null-point is equivalent (at 40°) to a temperature difference 

 of about '0041° 0., and the total loss or gain by radiation, &c, 

 corresponding to this difference between 6 X and Q would not 

 exceed 0*2 thermal gram per half-hour, whereas the actual 

 duration of the majority of these experiments was about 18 

 minutes. 



The corrected formula was, however, used throughout these 

 experiments for the adjustment of the contact-maker. 



* [Note by E. H. G., August 12, 1895.— I have found by re -standard- 

 izing the bridge-arms, that the above supposition was correct.] 



