CHEMISTRY — RICHARDS. 221 



namic computations. The new method of adiabatic calorimetry recently 

 developed at Harvard is especially suitable for cases of protracted reaction 

 of this kind. The study of these reactions led to the discovery of several 

 minor calorimetric errors which had vitiated most of the earlier determina- 

 tions; the sum of these errors amounted in some cases to a considerable 

 quantity. Especially it was shown that more thorough stirring than is 

 usually practiced is absolutely essential for great accuracy. The heats of 

 solution of zinc, iron, aluminum, cadmium, and magnesium were determined 

 with great accuracy, hydrochloric acid being used as a solvent. In order 

 that all the results might be calculated to the same standard, the heats of 

 dilution of the hydrochloric acid of the residual mixtures in the calorimeter 

 were in each case determined and the data all finally calculated so as to rep- 

 resent the heat of solution of i gram-atom of the metal in acid with 200 

 gram-molecules of water. Moreover, the heat of precipitation of silver from 

 a solution of silver perchlorate by cadmium was determined in order to serve 

 as a basis for the heat of formation of silver salts. 



III. Heats of Combustion of Certain Liquid Hydrocarbons: 



In the further prosecution of the revision of thermochemical data, the 

 heats of combustion of a number of octanes and zylenes were carried out 

 with unusual care. The object in choosing these substances was to endeavor 

 to trace the effect of constitution or arrangement upon the heats of formation 

 of isomeric substances and thus to obtain more definite ideas of the relation 

 of total energy change to structure. The hydrocarbons were prepared in a 

 state of great purity by Dr. Lathan Clarke according to methods worked out 

 several years ago for this purpose. Octanes in particular were chosen be- 

 cause the molecule is large enough to admit of considerable variety in the 

 isomeric compounds, but not so large as to confuse the relationships. Sim- 

 plicity in interpretation is also gained by having only two elements present. 

 In the execution of this work the adiabatic method of calorimetry already 

 mentioned in the previous paragraph was applied with equal success, and in 

 general the precautions used in previous work of this kind were adopted 

 throughout. 



Further investigation of the burning of these very volatile compounds 

 showed that special pains must be taken with regard to the form of the con- 

 taining vessel and the details of manipulation in order to insure complete 

 combustion. The volatile liquids were sealed in flexible glass bulbs and 

 ignited by means of small weighed quantities of sugar in such a way that 

 the combustion was in every case complete. The results showed very satis- 

 factory agreement among themselves. The details may be omitted here; 

 but it is interesting to note that the heats of combustion of the liquid isomers 

 were very nearly the same in every case. The maximum deviation amounted 



15— YB 



