rHi3iia).sE iJvOWER^. 327 



but even in this case it will be seen that the excess of rose- 

 centres only amounts to a very small percentage, and it 

 would almost certainly have been reduced, or perhaps eiFaced 

 altogether, if double the number of flowers had been ex- 

 amined on that day. 



The evenness in the general distribution of both forms 

 in this island, and their close approximation to absolute 

 equality in point of numbers, are certainly very remarkable. 

 In order to balance exactl}- each form should, of course, 

 have been represented by fifty per cent, of the entire number. 

 How very close an approach was made to this result will 

 be seen by the following figures. During the whole season 

 the total number of Primrose flowers which we examined 

 and noted amounted to 5,289, and they were divided in 

 this way : 



2,653 pin-centres = 50' 16 per cent. 

 2,636 rose-centres = 49"84 per cent. 



I have no reason to suspect that the examination of a 

 very much larger number of flowers — say ten times as 

 many — would have shown any substantial diflference in the 

 percentages, judging from the number of excursions we 

 made, and the varied character of the country which was 

 worked. We took in all sorts of places where Primroses 

 grow — lanes, cliffs, fields, streamsides, furze-brakes, sheltered 

 valleys and exposed moors — and never did I discover any- 

 where the slightest indication that one form preponderated 

 to any appreciable extent, nor could I perceive that situation, 

 aspect or season made any diflference. I found that the 

 sunny side and the shady side of the same lane yielded 

 on the same day a similar proportion of both forms ; and 

 that a warm sheltered valley in the interior of the island, 

 such for example as the one below Les Norgiots, at St. 

 Andrew's, with streamlets, trees and moist meadows, pro- 

 duced an assemblage of Primroses in which the two forms 

 were just as equally divided as on the exposed windswept 

 cliflP-slopes which face the sea at St. Martin's Point. 



Cases are well known among heterostyled ])lants in which 

 the flowers adapted for cross-fertilisation are larger and more 

 conspicuous than the others of the same species and growing 

 in the same locality that are normally autogamous. But it is 

 not so with the Primrose. And again, there are plants in 

 which the long-styled and the short-styled flowers develop 

 at different times, one kind being in t'ull bloom while the 

 other kind is still in bud. So one might have expected that 

 the rose-centred Primroses being primarily adapted for self- 



