ii DEVELOPMENT AND ENVIRONMENT 21 



balanced on a similar couple in a water-bath of known 

 temperature. 



Consequently, the two variables under examination 

 were recorded with unusually high precision, and relatively 

 rapid temperature changes could be employed with safety. 



The general arrangement of the initial work was 

 such as to ensure that temperature, and temperature 

 alone, should be the factor which limited the growth. 

 The rate of temperature-change was usually about 1 C. 

 in three minutes, so as partially to avoid errors from 

 prolonged exposure to high temperatures, and the air 

 space of the chamber was filled with pure oxygen. Under 

 these conditions growth curves like Fig. 31 were obtained 

 from fresh cultures, recently renewed, and kept at low 

 temperatures, so as to be free from any suspicion of 

 staleness. These curves indicated that growth followed 

 chemico-physical laws, being accelerated by increased 

 temperature, and approximately doubled in velocity by a 

 rise of 10 C. This acceleration began to fail in the 

 neighbourhood of 36 C., arid growth ceased altogether at 

 a mean temperature of 37*5 C., the maximum. A value 

 of 0'5 C. covered departures from normal in the 

 position of this maximum. Obviously, a second factor 

 was coming into play, antagonistic to the normal accelera- 

 tion, and probably itself accelerated by rise of temperature. 



Slower rates of heating, down to the maintenance of 

 culture-flasks at various constant high temperatures, 

 showed that the time of exposure had an influence on this 

 factor. The more rapid the heating, the nearer the 

 growth curve approached to its hypothetical form, i.e., a 

 sudden arrest at 37'5 C. when in full swing. The slower 

 the heating, the lower was the temperature at which the 

 slowing of the growth-rate appeared on the curve, and the 

 lower the temperature at which growth ceased. 



Following this graduated series of conditions down to 

 the stage at which the change of temperature is infinitely 



