367 



SOME REMARKS ON THE METHODS FORMULATED 



IN A RECENT ARTICLE ON ''THE QUANTITATIVE 



ANALYSIS OF PLANT GROWTH." 



By R. A. FISHER, 



Chief Statistician, Rothamsted Experimental Station. 



In the Annals of Applied Biology has appeared a paper (4) by Briggs, 

 Kidd and West; according to the authors "The series of articles of which 

 this is the first instalment, constitutes an attempt to formulate methods 

 for the quantitative analysis of plant growth and to apply these methods 

 to data which have been lying dormant in the literature for forty years." 



That the paper here criticised is primarily concerned with the 

 methods by which primary observations are to be treated in the study 

 of plant growth is emphasised by a reference made to it in an earlier 

 paper (3) in the Neiv Phyiologist. "The Relative Growth Rate, R, is the 

 weekly percentage rate at which the dry weight increases. It may be 

 assumed for purposes of calculation that the increase from week to 

 week takes place exponentially. R being the exponent, or that it takes 

 place linearly. Both are approximations. As to the relative merits of 

 the two different methods the reader is referred to (-1)." 



It will be noticed that no definition is here given of the Relative 

 Growth Rate, R, except for the case in which the mass increases exponen- 

 tially, when this is so 



m = m e Rt (I) 



where m is the mass and t the time, and R may be calculated from any 

 two observations: 



R = log m 2 - log m x 



t<i — ti 



where the suffixes 1 and 2 indicate the first and second observations. 



This is the quantity termed the Efficiency Index by Blackman(i), 

 and which may be correctly termed the Relative Growth Rate; we shall 

 use the latter term. No indication is given above as to what meaning is 

 to be attached to the term Relative Growth Rate when it is assumed that 

 the increase takes place linearly or as to how it is to be calculated from 

 the observations. 



