THE TIME OF ORGANIZED BEINGS I5I 



and that we rejoiced prematurely? It suffices to reflect on the 

 manner in which it is calculated to understand that the reason 

 for these fluctuations lies not so much in an error of reasoning 

 as in the method employed to calculate the values of i. Indeed, 

 these values are obtained from table, p. 85. Most of them 

 are interpolated, and even though the experimental values of i 

 are established on a fairly large number of experiments (about 

 six hundred), it is impossible to affirm that all these values 

 are rigorously exact. Ten thousand experiments would be 

 needed in order to consider the mean values as being really 

 representative. We stated above that the individual differences 

 due to the state of health, to the physiological age, were at first 

 glance negligible, with the exception of certain special cases 

 such as syphilis, diabetes, alcoholism. But it is probable that 

 they nevertheless exist, and that the indices drawn from the 

 chart (p. 87) are only the approximate expression of the true 

 index. Furthermore, differences can exist according to the 

 part of the body on which the wound is located, when the 

 granular contraction depending on the elasticity of the sub- 

 jacent conjunctive tissue is less efficient. For example, 

 wounds on the skuU, on the antero-internal part of the leg (on 

 the tibia), those on the foot or the back of the hand, are 

 characterized by a lower index, owing to the fact that the skin 

 lies almost directly on the bone. Not only is the contraction 

 weaker in these wounds, but blood irrigation through the 

 conjunctive tissue is diminished and can even be null if the 

 bone is bared. All these causes of error introduce small varia- 

 tions in the calculation, but can be easily taken into account by 

 slightly correcting the index if one wishes to establish the 

 curve of a particular wound. 



We can, moreover, assure ourselves that the variations of A 

 are due to fluctuations of a negligible magnitude in the values 

 of i. From a practical point of view, only the two first 

 characteristic figures of the values of i are significant, and it 

 can be said that a satisfactory mean accuracy corresponds to 

 values which are correct within ±0-002. Now, if we take the 

 value A which represents the largest divergence from the 



