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



correspondence between the increase in dry weight and the starch equivalent 

 of the gain of carbon to enable a good deal of wide comparative work with 

 varying plants and conditions to be done by the dry- weight method without 

 much vitiation by such errors of interpretation. 



For more exact work a convenient and rapid method of carbon analysis 

 would be a distinct advantage. The wet-combustion method described by 

 Hall, Miller, and Marmu,* depending on the absorption and estimation of 

 carbon dioxide by the method elaborated by Brown and Escombe,t has been 

 suggested to me by Dr. F. F. Blackman as possibly adaptable to the purpose. 

 Besides rapidity and convenience, it would have the advantage that the 

 whole of the carbon is driven off from the ash as well as from tlie organic 

 material, and ash determinations would be avoided. However, the possi- 

 bility of using this method has still to receive further consideration, especially 

 in respect to the small quantity of material which can be analysed, owing 

 to the limited amount of carbon dioxide that a Keiset tower will efficiently 

 absorb. It must suffice here to point out that along some such lines 

 modifications may be possible which would avoid all the errors of interpreta- 

 tion that have their origin in differences of composition of the photosynthetic 

 products. 



Section IV. — The Error froji Lack of Symmetry. 



This source of error, whicli Brown and Escombe conclude to be serious, 

 SachsJ apparently assumed to be negligible when dealing with corresponding 

 patches on either side of the midrib of a leaf. 



He realised, howevei-, the necessity of taking precautions to ensure as 

 great a degree of similarity as possible between the i)ieces of material to be 

 compared. Tlie same templets were used on the two sides of the k>af, and 

 care was taken to include as nearly as possible the same proportion of veins 

 in corresponding pieces, and where possible to avoid outstanding veins 

 altogether. The leaves were also carefully selected for. their freedom froni 

 bleinish(!s, such as (by spots or crumpled areas. 



Sachs also used a number of leaves for each experiment, and so reduced 

 the probable error from asymmetry, though he docs not make a point of this. 



Meiize, experimenting with comparatively small loaves, had to use tlie 

 gi'cater part of each half-leal', and could not select similarly veined portions. 



* "'I'lic Ivst itriatimi of ( 'arljon in Soils aii<l Kiiulred SubstaiiceH," '.Jouru. (!iioiii. Soc, 

 vol. H!>, "rraiiH.,' \'.m, p. m>. 



+ ' I'hil. TraiiH.,' li, vol. 1!)3, 1900, p. 'im. 

 \ Loc. cil. 



