674 TEMPERATURE AND -LATENT PERIOD 



in which e is the Naperian base, ^ is a quantity characteristic of a 

 given chemical reaction, and R is the gas constant which may be put 

 equal to 2. 



Our hypothesis states that the duration of the latent period de- 

 pends on the time required for the reaction L^T to form a certain 

 amount of the substance T. If this is correct, the relation between the 

 temperature and the latent period should be adequately expressed by 

 equation (3). Since this equation requires only the ratio between the 

 two velocity constants, it is permissible to substitute in their places 

 the reciprocals of the time required to perform a definite amount of 

 chemical work. The latent period is the time required for the for- 

 mation of a definite amount of the material T. Therefore the re- 

 ciprocal of the latent period at different temperatures may be used 

 in the equation of Arrhenius. The heavy line drawn in Fig. 2 is the 

 curve for equation (3) when ix = 19,680. In making the calcula- 

 tions it must be remembered that the latent period is the difference 

 between the observed reaction time and the sensitization period. 

 The average sensitization period for these temperatures, as already 

 stated, was found to be 0.21 second. 



The points in Fig. 2 are seen to be well represented by the theo- 

 retical curve from 13° to 21°. Above 2rC. the experimentally 

 determined values deviate decidedly and increasingly from the ex- 

 pectation according to the Arrhenius equation. The experiments 

 were performed at Woods Hole, Mass., during July and August, 1918. 

 These are the hottest months of the year. During 1918 the mean 

 water temperature for July was 20.2°C.; that for August was 21.7°C. 

 The highest water temperature recorded was 23.0°C. for July 30. 

 It is therefore apparent that the deviations of the experimental data 

 from theoretical expectation begin to occur at temperatures above 

 those to which the animal is normally subjected, even in the hottest 

 days of the year. Below this normal maximum of approximately 

 21°, the data follow accurately the expectation that the latent 

 period is conditioned by a single, simple chemical reaction. 



The increasingly greater difference at higher temperatures be- 

 tween the calculated curve and the experimental data indicates 

 clearly that a second factor has entered, the effect of which becomes 

 more and more patent as the temperature increases. If we assume 



