o28 MMMKOSi: KL<)Wi:i;S. 



pollination would have been more ])revalent clurino- the early 

 spring, which in these islands is usually bleak and cold, and 

 consequently unfavorable to the appearance of insect life ; 

 Avhilst the pin-centred Primroses, depending- as they must for 

 their fecundation upon the ministrations of small insects, 

 might be expected to blossom in inci'easing numbers as the 

 season advanced, so as to gradually replace the others. But 

 the observations I have recorded above clearly show that the 

 advance of the season has little to do with it, and does 

 not really in any degree affect the surprising uni ormity 

 which exists in the average numbers of the two forms in 

 this island. 



