INFLUENCE OF TEMPERATURE ON CHEMICAL REACTION 211 
It becomes, therefore, a matter of more than casual interest to the 
physiologist that by application of the principle of adiabatic calori- 
metry, recently developed by T. W. Richards into a method of remark- 
able accuracy, whereby quantities of heat heretofore unmeasurable 
may be readily determined, it is possible to extend our knowledge of 
internal energy changes to regions yet unexplored. Already, specific 
heats, heats of solution, of neutralization, of dilution, have been meas- 
ured with a precision impossible a decade ago. In like manner it has 
been shown that heats of combustion may be determined with sufficient 
accuracy to show measurable differences between the heats of forma- 
tion of isomeric substances, and thus to indicate for the first time 
something of the character of the correspondence between internal 
energy and molecular configuration.^ It is not impossible that by 
extension of this method the heat of slow-moving reactions like saponi- 
fications and hydrolyses (if in fact these yield quantities of heat com- 
parable with those set free in the dilution of salt solutions) may be 
experimentally determined. The results of such work, if positive, 
would indicate one way among many in which purely physical research 
may be of service to the biologist. 
The foregoing statements summarize in a general way the influence 
of a change of temperature upon heats of reaction ; and thus show the 
products of combustion were determined with the greatest care. The results of 
twelve rest experiments showed in the average a discrepancy between the heat energy 
released and that calculated from the heats of combustion of the materials used up 
of only 0.04 per cent.; in twenty work experiments the average discrepancy was 0.13 
per cent.; while that of the whole thirty-two measurements was 0.08 per cent. 
This degree of precision is comparable with that of the most accurate calorimetric 
determinations yet made (6). 
^ For instance, heats of neutralization have recently been measured by the 
adiabatic method with an experimental accuracy of two or three hundredths of a 
per cent., and the temperature coefficient of a reaction of this type with the remark- 
ably small probable error of one part in ten thousand. In the latter case, the heat 
of reaction at different temperatures calculated from the experimentally determined 
coefficient was checked by subsequent observation with this average precision, and 
with agreement in half of the recorded measurements of six parts in one hundred 
thousand. The specific heats of liquids, also, have been measured with an accuracy 
comparable with that reached in the determination of heats of neutralization. 
Similarly, the heats of combustion of the four isomeric hydrocarbons pseudocumene, 
mesitylene, normal and isopropyl benzene have been determined to be respectively 
1243.5, 1246.4, 1249.0, and 1250.0 kilogram calories, and subject to a probable error 
no greater than a tenth of the larger differences (7). 
