INJURY AND RECOVERY 121 



priate for each mixture, as given on page 140. In general 

 the agreement is satisfactory. 



With so large a number of constants it might seem 

 possible to fit any sort of curve, and hence the significance 

 of the actual accomplishment might be lessened. This, 

 however, is by no means the case. 



It should be noted that we do not employ new constants 

 to fit these curves, but that in every case we use the con- 

 stants already determined as the result of other and 

 quite different experiments. In view of this the results 

 have a special significance. 



If we accept the conclusions stated above we are 

 obliged to look upon recovery in a somewhat different 

 fashion from that which is customary. Recovery is usu- 

 ally regarded as due to the reversal of the reaction which 

 produces injury. The conception of the writer is funda- 

 mentally different; it assumes that the reactions involved 

 are irreversible (or practically so) and that injury and 

 recovery differ only in the relative speed at which certain 

 reactions take place. 



It would seem that these experiments, and those pre- 

 viously described, afford a sufficient test of the theory. 

 It has been found that the agreement between the calcu- 

 lated and observed values is satisfactory whenever a 

 sufficiently large number of readings are averaged in 

 arriving at the observed values. 



In the foregoing account many details are necessar- 

 ily omitted, owing to lack of space. These, however, are 

 not essential to the main purpose, which is to show how 

 the process of injury and recovery may be analyzed and 

 subjected to mathematical treatment. Starting with cer- 

 tain assumptions we have formulated equations by means 

 of which we can predict the behavior of the tissue. If the 



