610 



REPORT — 1891. 



TUESDAY, AUGUST 25. 

 The following Reports and Papers were read : — 

 1. Report on Isomeric Naplithalene Derivatives. See Reports, p. 265. 



2. Report on Wave-Length Tables of the Spectra of the Elements. 

 See Reports, p. 161. 



3. Report on the AhsorjHion Spectra of Fure Compounds. 

 See Reports, p. 275. 



4. On the Specific Heat of Basalt. By W. C. Roberts-Atjsten, C.B., 

 F.R.S., and A. W. Rucker, F.R.8. 



Having been asked by the Rev. O. Fisher to determine for him the latent heat 

 of basalt, we made som.e experiments on a specimen which was furnished to us by 

 Professor Judd. Fragments of the rock were melted in a platinum crucible, the 

 junction of a thermal couple consisting of platinum with platinum containing 

 10 per cent, of rhodium was immersed in the pasty mass, which was then allowed 

 to cool. The scale of the galvanometer had previously been standardised by an 

 observation on the solidifying point of pure gold, and tliis was repeated from time 

 to time whilst the experiments were in progress. When the spot of light had 

 reached the desired point, the wires were nipped off close to the basalt, and the 

 crucible and its contents were plunged into 1,000 grammes of water contained in a 

 silver calorimeter. The water was stirred by a screw or fan of silver, which was 

 rotated by an electro-motor. The temperature was read by means of a mercurial 

 thermometer which had been carefully corrected. 



The two main sources of error in the experiments are probably an uncertainty 

 as to the mean temperature of the basaltic mass, O'wing to its being a bad con- 

 ductor of heat, and the fact that in the processes of heating and cooling, it 

 undergoes more or less important changes of constitution. 



The iirst error was reduced to small proportions by using small quantities ot 

 basalt, the most employed rarely much exceeding 20 grammes. 



The second error is iu part unavoidable ; the rapidly cooled basalt was always 

 glazed like olivine. AVe also found that frequent heatings and coolings, and the 

 nature of the flame, whether oxidising or reducing, employed to heat the mass, 

 appeared to atl'ect the results very seriously. 



In some experiments the crucible was heated in a small gas furnace, in others 

 in a coke furnace. All the former were consistent with each other, and those of 

 the latter group in which fresh specimens of basalt were used, were in agreement 

 with them. The results obtained with specimens which had been heated two or 

 more times in the coke furnace were, however, very irregular, and as we have not , 

 definitely proved what was the cause of these discrepancies, we publish our results 

 with a certain amount of reserve. In the following table T is the temperature (Cent.) 

 of the basalt at the moment of immersion, C is the mean specific heat between 

 about 20° C. and T :— 



