50 



On MendeVs Categories 



from Sandford in a straight line. Tlie second race was collected, and in large part 

 examined by Miss C. B. Sandei-s, whoni I giadly take this opportiinity of thanking 

 for the time and labonr she was kind enough to give. The epidermis was removed 

 frora the Icaf, and mounted in water, with its hair}' side upperinost, the hairs being 

 then counted under a low power of the niicroscope ; in this way error due to the 

 accidental breaking of hairs was avoided, because the stump of a broken hair was 

 easily recognised. 



TABLE I. 



Number of Hairs per Square centimetre of Lower Leaf in Lychnis diurna ? . 

 The area chosen in all cases so as to exclude the great veins. 



The figures given do not coiivey an adequate iinpression of the ränge of varia- 

 bility throughout the species. Eveu in iocalities close to those inhabited by the 

 races described a greater ränge of Variation has been observed ; such as they are, 

 however, the two races exaniiued are enough to show how wide is tl)e range, 

 covered by the category "hairy," in L. diurna alone. In a single race of Z. vesper- 

 tina fron) Sliotover Hill, near Oxford, it is even greater; 112 plauts of this race 

 included one glabrous individual, and one with 1106 hairs per Square centimetre 

 on the Standard leaf The distribution of hairs in tliis race is sfiven in Table II, 

 which is again an inadcqnate pieture of the ränge throughout the species, even in 

 the neighbourhood of Oxford. Odd plants gathered by the wayside have given 

 over 1300 hairs per Square centimetre. 



