M. J. Thomseir's Thermo-chemical Researches. 411 



too high, whilst that which they have obtained for sulphuric acid 

 is nearly exact. 



The experimental method employed was that of mixtures, the 

 calorimeter being carefully preserved from all external radiation. 

 The temperatures were determined by means of thermometers 

 calibrated and compared with the greatest care, and on which the 

 2 Jq of a degree could be read off with certainty. The weights 

 of the liquid mixed were usually 450 grammes of each of the so- 

 lutions. In the majority of cases these solutions contained 200 

 equivalents of water to one of the reacting body; sometimes, 

 however, solutions two or three times as dilute had to be em- 

 ployed. 



For solutions thus dilute we can, without fear of an appreci- 

 able error in the results, calculate the thermic effects by merely 

 taking into consideration the weight of the water. Moreover, as 

 the temperature soon acquires a state of equilibrium (in less 

 than a minute), whenever the variation in temperature does not 

 exceed a degree, it can be proved that it remains for several 

 minutes without appreciable change, so that there is no occasion 

 for any correction in the result obtained. In the few cases in 

 which this variation is more considerable, the thermometer must 

 be read every minute from the moment at which the mixture 

 was made ; then it can be calculated what the maximum tempe- 

 rature must have been, independently of external influences. 



The value in water of the vessel used as the calorimeter, and 

 of the thermometer, had been determined by experiment; it was 

 either 13 grammes or 97, according as one or the other of the 

 two vessels intended for these experiments was used. 



The exactitude of the method was established by numerous 

 experiments, in which water of different temperatures was mixed. 

 The maxima of error in forty-two experiments were + 7 and 

 — 8 thermal units. In a series of eighty- eight comparative 

 experiments, the departure from the mean only once reached 9 

 thermal units, and in eighty-four of these series it was less than 5. 

 Hence we may assume that the possible maximum of error in each 

 determination does not exceed +5 thermal units; and as the 

 quantity of matter employed in each of them represents in ge- 

 neral from 5 to ^ of an equivalent, the limit of the maximum of 

 error must be ± 20 or +30 thermal units for the thermic effects 

 proportional to an equivalent of the substance (the equivalent 

 of hydrogen being taken as unity) . 



The following are the results of these determinations, of which 

 we only give a summary. 



Neutralization of soda by sulphuric acid, nitric acid, and hydro- 

 chloric acid. — Both the base and the acid were in every case dis- 

 solved in 200 equivalents of water. The heats disengaged for 



