JOHN H. NORTHROP 267 



trypsin that will be inactivated in any given interval of time can be 

 approximated. The total amount of trypsin remaining in solution 

 can then be found by difference and the amount of this which will 

 be free and active at the concentration used in the determination, 

 and in the presence of the known concentration of inhibitor can be 

 calculated. This value should agree with that found by experiment. 

 It must be remembered that the above calculation is only a first 

 approximation since it contains two assumptions that are not strictly 

 correct. (1) That the free trypsin is inactivated at the same rate as 

 the same concentration of pure tr)^sin. As a matter of fact the 

 amount of trypsin destroyed under the conditions of the experiment 

 will be slightly greater than the quantity which would be destroyed 

 if there were no combined stable trypsin present, since some of this 

 will be dissociated as the free trypsin is destroyed and the amount 

 of free trypsin and therefore the amount destroyed increased in this 

 way. If the experiment is limited to the first part of the reaction 

 (as was done) this difference is within the experimental error. (2) It 

 was assumed that the only inhibiting (protective) substance present 

 was added as the inhibiting solution. The decomposition curve of 

 the trypsin solutions alone, however, show that in general there is 

 always some protective substance present in the trypsin solution. 

 The fact that the addition of inactivated trypsin renders the enzyme 

 more stable indicates also that the solution contains some protective 

 substances. This effect can hardly be ascribed to the inactive trypsin 

 itself since it was shown in the first part of this paper that inactivated 

 trypsin took no part in the reaction. The neglect of this quantity 

 tends to make the calculated amount of trypsin destroyed too low. 

 This is the result that is obtained. (If this quantity of inhibitor 

 present in the trypsin solution is taken into account, the calculated 

 and observed values may be made identical, but since there is no 

 independent method of determining the value to be used, the process 

 really consists in adding another arbitrary constant to the formula 

 and so does not add much to the validity of the proof.) 



The result of an experiment calculated and carried out as described 

 above is shown in Table II. The agreement is as good as could be 

 expected in view of the many sources of experimental error and of 

 the fact that the calculation involves the extrapolation of the cquilib- 



