28 LIBERATION OF ENERGY 



experiments which confirmed Crawford's results, and made firm 

 the principle of indirect calorimetry. They determined the 

 amount of ice melted by the combustion of a weighed amoimt of 

 carbon (a candle) and the volume of the COg evolved. A similar 

 experiment was then tried with a guinea-pig. They found that 

 for equal volumes of COg formed, the candle yielded 25-4 cals. 

 as against the guinea-pig's 31-8. The experiment is bristling 

 with errors, many of which the authors realised. For instance, 

 the respiratory and calorimetric determinations were not, as by 

 Crawford, made simultaneously, and obviously thermal condi- 

 tions were not the same. As we shall see later, cold raises the 

 CO2 output. If allowance is made for this and for other minor 

 errors, the figures for candle and animal come close enough to 

 justify the conclusion that the processes are similar, and that 

 the source of heat in both is the combination of C and O^. 



The various sources of error due to faulty technique have been 

 gradually eliminated, and the resultant calorimeters that bear 

 the names of Atwater, Rosa, and Benedict and that of Williams 

 produce results that are sufficient to convince even the most 

 sceptical of honest observers that the oxidation of assimilated 

 foodstuffs in the living body produces the same evolution of 

 energy as they would if burned in the bomb calorimeter, provided 

 the end-products are identical. 



The direct method is not of such general use as the indirect. 

 Study of the papers from the Carnegie Institute of Washington 

 or of those from Cornell University makes clear the complexity 

 of the machine and the intricacy of its manipulation. The cost, 

 except for the smallest Williams' boxes, is prohibitive. The 

 apparatus can be much simplified if the direct estimation of the 

 energy-changes is omitted and the observer confines himself to 

 measuring the respiratory gases and the urinary output. 



(b) Indirect. As we have seen, the basis of this method also was 

 laid by Crawford. It depends upon the following established facts : 



(I) The quantity of energy liberated depends on the chemical 

 composition of the food used. 



(II) The quantity of oxygen absorbed depends also on the 

 chemical composition of the food used ; therefore, 



(III) There must be some relation between the energy evolved 

 and the quantity of oxygen absorbed. 



(IV) The three proximate principles of food differ markedly 

 in chemical composition, (a) Proteins contain nitrogen, which 

 is eliminated almost entirely in the urine, (b) Carbohydrates and 

 fats differ widely in their oroportion of O to C. 



(V) If a determination were made of the amount of heat and 



