326 W. A. CANNON 



of growth inureased for 18 hours, more or less, and then de- 

 creased. In the specimen with the longest roots, approximately 

 the maximum rate was attained within 6 hours after the be- 

 ginning of the experiment, after which the rate was fairly uni- 

 form for about 30 hours and then declined. On the whole 

 the curves of root growth in this experiment resemble the curves 

 for growth of the longer roots in the previous experiment. 



Root growth at a soil tefnperature of 31. 5° C. Four series were 

 run at a temperature of 31.5°C. In these experiments the roots 

 of Nos. VII and XI had an initial length between 10 and 18 

 mm., those of No. VIII a length between 12 and 32 mm., and 

 those of No. X an initial length of 85 to 105 mm. The results 

 in experiments here numbered VII and X, which are to be con- 

 sidered representative, are given in tables 2 and 3. 



In the experiments where the initial root length is 32 mm. or 

 less, it was observed that usually there was a well marked in- 

 crease in the growth rate with the passage of time, and that the 

 minor fluctuations in the rate of growth were, as a rule, very 

 pronounced. The generalizations can also be made in regard 

 to root growth of such plants as had roots that were from 85 

 to 105 mm. in length at the beginning of the experiment, that 

 the minor variations in growth were as a rule not considerable, 

 and also that the curve of growth did not show the growth 

 increase with the elapse of time, that was observed in shorter 

 roots. On the contrary the growth rate remained approximate- 

 ly constant or gradually became less. As a corollary to this 

 it is of interest to note that the length of the roots in the former 

 instance — that is where the initial length was shortest — was 

 relatively great at the time of the most rapid growth. For 

 example, in no. VII the root lengths were 38.3 to 63.2 mm. in 

 length at this time; in the other experiments the root length at 

 the time was variable, but usually they were relatively long. 

 An examination of the notes accompanying the experiments 

 shows that the variation is not to be associated with the pas- 

 sage of time, the possible fluctuations in temperature of the 

 soil or of the possible variation in the water relations, and prob- 

 ably not with that of illumination. On the other hand it ap- 



