430 Mr. D. Thoday. Experimental Researches on [Mar. 1, 



leaves at the same time ; but, aa a consequence of this loss of sensitiveness, any consider- 

 able relative change in the readings becomes still more significant. 



(ii) Discussion of the Results obtained with H. annuus. 

 In the following table the results of all the experiments are collected 

 together. The leaves have been separated into five classes, according to the 

 degree of turgidity which they exhibited, and the results are so arranged 

 that the rates of increase shown by leaves belonging to the same class in 

 the different experiments may be compared. In the last column is given 

 for each class the average rate of increase of all the leaves included in it. 



Table VI. — All the figures are rates of increase of dry weight in milligrammes 



per sq. decimetre per hour. 



Condition of leaves. 



Exp. 1. 



Exp. 2. 



Exp. 3. 



Exp. 4. 



Exp. 5. 



Average rate of 

 increase. 



(1) Turgid 



17 -2 



17 -0 



17-1 





16 -6 



16 -9 





13 -8 









19 -1 







16 -3 









18 -6 















(9 -8*) 



(16 '1 including *) 



(2) Moderately turgid, occa- 





13-0 





15 -5 



9-7 



12 -5 



sionally rather limp 









14 -1 



10 -5 





(3) Limp 



9-9 



7-9 





8-0 





8-5 











8-5 







(4) Limp to flaccid 







.5 -3 







5-3 









4-0 















4-3 















7-8 









(5) Quite flaccid from begin- 



3-5 



-0-8 





4-1 





1-6 



ning of experiment 









-0-5 







The agreement of these experiments among themselves is very satis- 

 factory. The fact that the turgid leaves in Expt. 5 have given slightly 

 higher results on the whole than those in Expt. 1, is, I think, a significant 

 difference which is to be attributed to the different conditions under which 

 the two experiments were conducted. Considering such dilferences in 

 external conditions, and the range of variation possible in the condition of 

 leaves in the same class, a small amount of variation beyond that which is 

 due to asymmetry errors is to be expected in the individual results. The 

 average rate of increase for the class is thus the mean of individual rates 

 which vary within fairly narrow limits. 



Sachs, who used largo areas from seven leaves, obtained as the rate of 

 increase in the dry weight of detached, leaves in continuous sunshine 



