AN APPARENT SELECTION OF FORMS OF H. NEMORALIS 

 BY ADVERSE CONDITIONS. 



By a. E. boycott. 



(Read before the Society, March 12th, 1913). 



In 1907 I collected 310 specimens of a dwarfed (average i8"4 x i4'2 

 mm.) form of 7iemoralis from a sand-blown area of grass by the shore 

 near Land's End. As I was informed, the sand had been blown on 

 to, and more or less covered, the grass only within the preceding two 

 or three years. At the same time I took 79 specimens of larger 

 size (average 2 2 "2 x i8"r mm.) from among grass just beyond the 

 reach of the sand and within a couple of hundred yards of the area 

 where the smaller specimens occurred, x^part from the presence of 

 sand, the two areas appeared to be of exactly the same character. 

 The proportions of the two colour forms which were present and of 

 the different bandings varied a o;reat deal in the two series. 



Total number collected 

 libellula per cent, 

 opooo 

 00300 



12345 

 rubella per cent. 

 00000 

 00300 

 12345 



Grass 



79 



43% 



21 



26 



38 



57% 

 67 

 1 1 

 18 



Sand 

 310 



72% 

 0-5 



6 



77 

 28°/ 



78 



In the sand specimens, therefore, libellula were proportionately 

 more abundant (and the difference is beyond that which might be 

 attributed to errors of sampling) and the band formulae 00000 and 

 00300 were much less numerous. Altogether in the grass series 37 

 out of the 79 (or 47%) were 00000 and only one of the 310 sand speci- 

 mens. I am inclined to attribute the difference to the elimination of 

 the few-banded forms by the adverse influence of the sand. It seems 

 clear from the small size of the shells that the general conditions were 

 not favourable so that there is less to be said in favour of the alterna- 

 tive hypothesis that the sand encouraged the many-banded forms. 

 Among the sand specimens were 25 of the formula 00045 which did 

 not occur in the grass series. 



