8 Mr. D. Thoday. Experimental Researches on [June 11, 



Table V. — Hclianthus tuhcrosus : area by disc method. 



Assimilation experiment ; August 19, 1907. Leaves detached, in greenhouse, 

 under canvas. Time of exposure, 11.45 a.m. to 5.45 p.m. Intermittent 

 sun. Temperature 23° C. 



Number of 

 discs. 



Dry weight. 



Gain of dry 

 weight. 



CO2 found on 

 combustion. 



• 



Grain of 

 CO,. 



Starch 

 corresponding 

 to gain of COj. 



65 



{a) 0-1701 

 (5) 0-1923 



0 0222 



0-2689 

 0 -3020 



0 -0331 



0 0203 



In Table VI two examples are given, in which asymmetry in dry weight 

 and carbon content have been compared, to indicate the degree of accuracy 

 obtained and the kind of experimental error to be allowed for in considering 

 the results already given. 



Table VI. — Cherry Laurel : Examples of Asymmetry. 

 All numbers are grammes, and refer to the actual area used. 

 a = left half-leaf. h = right half-leaf. 





Dry weight. 



Ash. 



■ 



Dry weight, 

 less ash. 



Percentage 

 difEerence in 

 organic 

 substance. 



COo found 

 on combustion. 



Percentage 

 difference in 

 CO2. 



1 



(a) 0-1044 

 (i) 0-1024 



0 -0051 

 0 -0054 



0 -0993 

 0 -0970 



2-3 



0 -1927 

 0 -1868 



3-2 



2 



(a) 0-1541 

 lb) 0-1565 



0 -0080 

 0 -0074 



0 -1461 

 0 -1491 



2 -1 



0 -2783 

 0-2853 



2-0 



The differences in these percentages are of similar magnitude to the 

 discrepancies in the previous tables, and it appears, therefore, that on the 

 whole there is little in the results given in those tables to be explained 

 otherwise than Ijy errors incident to the teclinique of experiment. What 

 differences tlienj are, moreover, do not al'Ibrd any explanation whatever of 

 tlie large positive error found l)y JJrown and Escombe in results obtained by 

 the dry-weight method. The dillerences (ire small, and are for the most 

 part in favour of the apparent gain of carbon. Indeed, in Tables I to III 

 as a whole, the observcid increase in dry weight is less than that calculated 

 from the gain of carbon and ash. This might be due to the accumulation 

 of suiwtances containing a high(n- percentage of carbon than starcli ; or to 



