1909.] 



Vegetable Assimilation and Respiration. 



3 



Section II.— On the Nature of the Errors to which the Half-leaf 

 Dry- Weight Method is Liable. 



The Sachs method consists essentially in the comparison of the dry weight 

 per unit area of one half of a leaf (the control half-leaf) at the beginning of 

 an experiment, with that of the other (the experimental half-leaf) after a 

 certain number of hours of assimilation. The difference is taken to be the 

 weight of the products which have been accumulated by the leaf during that 

 time per unit of area. 



Underlying the method are two assumptions, (1) that the symmetry is 

 perfect, i.e. that the two halves of every leaf used possess appreciably 

 the same dry weight per unit area, and (2) that no change takes place in 

 the experimental half leaf to alter the original dry weight per unit area 

 except tlie accumulation of the products of photosynthesis which it is the 

 object of the experiment to measure. 



The few experiments of Brown and Escombe clearly prove these assump- 

 tions to be unjustifiable, as leaves are far from symmetrical, and leaf area 

 may change under experimental conditions. Further, to help explain the 

 recorded excessive gain of dry weight during assimilation, they suggest 

 a possible increase in the retention of water by colloids of the insolated leaf.* 



If increase of dry weight were partly due to such an indeterminable 

 fixation of water, the increase could not be used as a measure of assimilation, 

 and the whole procedure would fall to the ground. This fear will be shown 

 in Section III to be without foundation. 



Catalpa hignonioides, used by Blown and Escombe in their test of the dry- 

 weight method, happens to have been an unfortunate choice, since its leaves 

 are particularly unsymmetrical in respect of dry weight. Other leaves show 

 a much closer agreement between the two halves. The degree of asymmetry 

 of a number of species is dealt with in Section IV. 



Change of area has proved to be of extreme importance, as it is to errors 

 from this source that the observed tendency to high resvdts is due. 

 Section V is devoted to this question. 



In later sections of the paper minor errors of technique are carefully 

 considered and estimated. Errors in the measurement of area, errors in 

 weighing due to incomplete drying of the hygroscopic leaf material, etc., 

 become of real importance, because their effects are cumulative in the 

 resulting error in the observed gain of dry weight. 



The following is a formal analysis of the various sources of error. Of 

 those classed as errors of interpretation, that due to the varying composition 



* Loc. cit. p. 59, 



B 2 



