RICHARDS AND ROWE. — SPECIFIC HEATS OF LIQUIDS. 195 



The mean value, then, for the specific heat of HCI.IOOH2O over 

 the range between 15.9° and 20.1° is 0.9634 compared with water 

 over the same range of temperature. This is equal to the results 

 found in the preliminary series, within the range of the probable error 

 (0.0001). It will be noted that the maximum difference from the 

 mean is considerably less than 0.1 percent. The probable error of 

 the seven determinations is only about 1 in the fourth decimal place. 

 These determinations are typical of the experiments made with the 

 other solutions. Because of the great similarity of the experiments • 

 to one another, it will not be necessary to record every experiment, 

 but the average values similar to the total averages given on the last 

 line of the table above will be given for each of the essential data, 

 together with the number of experiments. 



In the course of the experiments first performed, namely, all those 

 on hydrochloric acid and sodium hydroxide, the original preparation 

 of sulphinic acid was exhausted. This solution (designated A) 

 contained, as has been already stated, 6.275 percent of sulphuric acid. 

 For the remaining experiments it was necessary to prepare a new 

 solution (designated B below) which was found to contain 6.283 

 percent of acid.-"^^ In order to make assurance doubly sure, this new 

 solution was used in ten experiments with pure water in the calori- 

 meter which are recorded at the foot of the table below. The slightly 

 greater strength of the acid caused the average rise of temperature to 

 be slightly greater in this last series than in the first. At 20. 19° the 

 original acid would have caused a rise of 4.229°, and with this new 

 acid, 0. 13 percent more concentrated, the rise was 4.2.34° — an exact 

 proportionality. No more satisfactory evidence of the consistency of 

 the work could have been desired. 



12 In this connection a somewhat interesting demonstration of the accuracy 

 of the method was obtained. The sulphuric acid was originally analyzed 

 by titration with a sodium hydroxide solution which had been carefully stan- 

 dardized but which had stood for some months in a paraffin-lined bottle. The 

 mean of the analyses thus conducted, gave the strength as 6.272 percent. 

 When the heat evolved during the neutralization of § molecule of acid was 

 determined and compared with that shown by the previous lot of acid, the ob- 

 served and calculated results showed a serious discrepancy. The strength of 

 the new acid was calculated from the thermal data and found to be 6.282 

 percent. Investigation of the bottle containing the standard alkali showed a 

 break in the paraffin lining which had permitted contact between the glass and 

 the solution; hence the alkali had become slightly more alkaline; a fresh alkali 

 solution was prepared and carefully standardized and by its use the acid was 

 again analyzed. The result of these analyses is the figure given above, namely, 

 6.283 percent or within 0.02 percent of that calculation from the thermal data. 

 It is clear, therefore, that the calorimetric process had been so effectively ar- 

 ranged as to constitute a very accurate process of quantitative acidimetric 

 analysis. 



