WALLACE O. FENN 333 



out that one should compare the times necessary for equal amounts of 

 action rather than the amount of action at equal times. This is 

 equivalent to saying that the rates of two reactions should be com- 

 pared at corresponding stages. In the curves of Fig. 1, however, the 

 corresponding stages are not points of equal amount of action because 

 the maximum varies at different temperatures, but, rather, they are 

 points of equal percentages of the total amount of action possible at 

 that temperature. For comparative rates, therefore, we may take 

 the number of bacteria ingested per leucocyte per minute during the first 

 half of the reaction; i.e., until one-half the maximum number of bacteria 

 has been ingested. This criterion yields a value for the rate of the 

 reaction which is far from being ideal but which seems to be the best 

 approximation possible under the circumstances and certainly more 

 rational than the original. 



Following this procedure the rates of the reactions at different 

 temperatures have been calculated from the data of Madsen and 

 Watabiki.- In order to calculate from them the temperature co- 

 efficient, Qio, of the reaction the logarithms of these rates have been 

 plotted in Figs. 2 and 3 against the corresponding temperatures. The 

 temperature coefficient for any interval of 10 degrees on the abscissae 

 is the antilog of the difference between the ordinates at the two 

 temperatures; i.e., the slope of the graph for that interval. The 

 resulting graphs are practically straight lines which is rather an 

 unusual result for biological processes. This means that the tem- 

 perature coefficient is constant over the entire range from 5°-35°C. 

 ^10 was found to be 2.05 ±5 per cent and 2.0 ± 5 per cent in Figs. "2 

 and 3 respectively. For comparison with these curves the values of K, 

 calculated by Madsen and Watabiki according to the formula for a 

 monomolecular reaction, are also plotted in Figs. 2 and 3. Our 

 improved analysis evidently gives a smoother curve, the probable 

 error being only about one-half as large. 



In a recent paper on phagocytosis (2) a formula was derived for 

 calculating the chances of collision between leucocytes and particles 

 of known size and density when stirred together in a common suspen- 



2 Madsen and Watabiki (1), Tables 1 and 2. The data on other tables could not 

 be similarly treated because the experiments were not carried to completion. 



