CLASSIFICATION OF THE PRIMORDIA 69 



are given for certain species but not for others. We are in- 

 formed that the leaves are very narrow at their base in certain 

 species and narrow or broad in others, but for many species 

 the property breadth at the base is not mentioned. Here three 

 properties may be measured : (i) length ; (2) greatest breadth ; 

 (3) breadth at the base. I have carried out this work for ten 

 British species of the genus Mnium,^ replacing the terms long, 

 short, large, small, broad, narrow, very narrow, etc., by three 

 figures for each species. This subject is rather complicated 

 because gradation comes into play here. (See Part VIII.) 



§ 53.— CLASSIFICATION OF THE PRIMORDIA.— In the 

 present work a number of primordia have already been men- 

 tioned — for instance, the length, the breadth and the number 

 of cells of certain organs ; the dimensions of the cells ; the 

 presence or absence of hairs ; the number of hairs on the 

 labrum of an insect ; certain colours ; the number of seeds in 

 a pod, etc. Between those properties no relations exist ; they 

 have been taken at random, as examples. When we consult 

 a flora or a fauna in order to find primordia (§52) we are pro- 

 ceeding at a venture. We behave as if an indefinite number 

 of properties (so-called characters) existed in each species ; in 

 each pecuUar case we pick out and investigate some of them, 

 but the choice we make is arbitrary. Similar remarks are 

 applicable on the mass of different properties which have been 

 investigated by the Mendelian School and by the biometricians. 



We need a method enabling us to compile an inventory of the 

 primordia of each species : such a method ought to be based 

 upon a classification of the primordia. In § 46 I have attempted 

 to classify the simple properties (primordia) into nine groups 

 according to their development. This classification may 

 render service in embryology but is inadequate when we want 

 to make up a hst of the properties of a species. 



W. BATESON 2 has divided the phenomena of variation 

 into two groups : variations in the processes of division and 

 variations in the nature of the substances divided [loc cit. 

 chapter ii., p. 31). This classification of the variations is in 

 reality equivalent to what I call a classification of the primordia. 

 BATESON calls meristic variations the variations in the pro- 

 cesses of division and substantive variations the changes in 

 actual composition of material. He also uses the terms 



^ See my paper : " Quantitative Description of Ten British Species of the 

 genus Mnium. Linnaan Society's Journal — Botany, vol. xUv., November, 

 191 7. (Communicated by Prof. F. E. V^EISS, D.Sc, F.L.S. Read ist June 

 1916.) 58 pages, 9 text-figures. 



* W. BATESON, Problems oj Genetics (New Haven, Yale University Press. 

 London, Humphrey Milford. Oxford, University Press). 1913- 



