HEAT EQUIVALENT OF NUTRIENTS OF FOOD. 593 



fact that tlie couibustiou was not a direct one, but was accompanied by 

 a number of the secondary reactions, some of which iutinenced the 

 results on the positive and some on the negative side. These errors 

 could only be estimated by indirect methods, for at the time these 

 investigations were made there was not a single substance available 

 whose heat of combustion had been determined by reliable methods. 

 Had we then possessed a substance with a known heat of combus- 

 tion, so that a definite amount of heat could be produced at will, it 

 would have been an easy matter to determine the absolute value of 

 these secondary thermal processes. In the determinations, however, 

 the constants representing these secondary thermal processes had to 

 be learned by indirect and tedious methods, and as a matter of course 

 each error in the constants affected the final results. 



In the course of time the discovery of Berthelot's bomb made pos- 

 sible great progress in thermochemistry, since it is an instrument whicli 

 measures the heat of combustion of all combustible compounds with 

 astonishing accuracy. 



We have used this apparatus since 1887.^ I have described it in 

 detail in a former publication.^ 



We have repeated with it nearly all the exf)eriments made by the 

 Ijotassium chlorate method. This repetition has shown that in these 

 earlier results there is a constant error, although these results agree 

 very nearly with each other. This error is seldom over 2 per cent, 

 often less, and is always on the same side. If this amount is added 

 to the results obtained by us with the potassium chlorate method, 

 values are obtained which are accurate enough to meet the present 

 requirements. 



This error, which in many cases is of very little moment, is due to 

 the fact that iu spite of all efibrt it was not possible, with the indirect 

 method then available, to estimate accurately the value of the constants 

 which represent the secondary thermal processes in the combustion. 



I think I owe it to myself and those who have worked with me, since 

 we have worked so earnestly, to make these explanations here, and it 

 is also desirable to do this since the method we had followed has been 

 the subject of uiuch unjust criticism. The i^otassium chlorate method 

 was abandoned, not because with the improvements I had made it was 

 inaccurate, but because the Berth elot and Mahler apparatus was much 



• Berthelot's apparatus consists of a steel bomb, cylhidric.al in form and lined with 

 platinum. The substance to be burned is held iu a platinum crucible inside the 

 bomb. Oxygen is forced into the apparatus until the pressure is 20 or 25 atmos- 

 pheres. The bomb is immersed in water and the substance is ignited by an electric 

 current. The rise in temperature of the water shows the amount of heat produced 

 by the combustion of the substance. The apparatus is described in Bulletin No. 

 21 of the Office of Experiment Stations on Methods and Results of Food Investi- 

 gation. 



^ Jour. prak. Chem., 2d ser., 39, p. 503. 



