W. F. I{. Wki-don 53 



The foregoing facts sliowtlu: laiitfc of vaii:itiön in thc nuinber dl' hair.s per unit 

 area cm tlic leaves of a fcw iiidiviiliials. Il is dilHiuilt. io estiinate the percentage 

 freipiciicy of irlaiuliilar hair.s, hecaiise tliey are inore trequent near the base of the 

 leaf tliaii near its apox; I have atteinpted an estiinate of their ))ercentage freijuency 

 in the basal part ot tlie leaf, and the result is given in l'able IV. 



The few data here brought together are sufficient to sliow tlie way in which 

 the adoption of such a category as "hairy" eonceals the facts of Variation within 

 the races disciissed. In the hght of such facts the statenieiits tiiade by Mr Bateson 

 and Miss Saunders are seen to be utteriy inadäquate, either as a description of 

 their own ex])eriinents, or as a (K-nionstratioo of Mendel's or of any other laws. 



Whcn hairy and giabrous plants were crossed \ve are told that "aniong the whole 

 numbcr of jjhvnts raiscd, not a singlc interniediate was observed" ( I.e. p. 1 ö), but we 

 are not told what "an intermediate" is. The authors nuist know, or ihey could 

 not certity its absence ; it would have been well if they had thought tit to define 

 an " intermediate " in their Report, for the definition is a vital part of their 

 argument. 



The prececling tables show that " hairiness " is not an absolute, invariable 

 quality, but that it is manifested in various degrees. In the few individuals 

 exaniined it is possible to pass by a series of small steps from the giabrous 

 condition through individuals with various nunibers of hairs per Square centi- 

 metre of leaf-surface, up to a condition of very great haiiiness. It is perfectly 

 legitimate to regard those individuals with a small number of hairs per unit area 

 as intermediate between those with a larger number and those with a smaller 

 number or with none. Thus a plant with only 80 hairs per unit area of leaf- 

 surface may be called intermediate between a plant with 200 hairs per unit area 

 and a plant with none. 



Other conditions are conceivable, intermediate between that uf a plant with a 

 given number of hairs per unit area, the hairs being of kuowai length, and that of 

 a giabrous plant. Thus the transition might take place by a reduction in the 

 length of the hairs through various steps to zero, without reduction in their 

 number ; in which case the plant with the shorter hairs would be intermediate 

 between the plant with longer hairs and the giabrous individual ; or again the 

 transition might conceivably be etfected by the appearance of giabrous patches, so 

 that the intermediate individuals assumed a mosaic character. On the whole, 

 however, the density of the hairs per unit area seems the best measure of "hairiness" 

 for our preseut purpose. 



Whatever we may choose as a measure of hairiness, Mr Bateson and Miss 

 Saunders give no evidence by which we can judge the result of their work. Thus 

 we are told that "haiiiness" is dominant, becansc "of the thousand cross-breds 

 " raised from various unions between hairy and giabrous strains, all, without e.xcep- 

 " tion, were hairy" {I.e. p. 19); but unless we know how hairy they were, we cannot 

 judge what is the valuo of the Statement that hairiness is dominant. Thus a plant 



