X 
REPRODUCTIVE ORGANS IN THE PRIMULACER. 117 
its hybrid unions with P. Scotica. From these results of the 
short-styled hybrid unions, I naturally anticipated, from the 
evident structural affinities of the two species, somewhat similar 
results from the long-styled unions. In this, however, I was com- 
pletely wrong; for, after a number of careful experiments, I have 
failed to get a single seed from the long-styled P. farinosa by 
pollen of P. Scotica. How clearly do such cases show us that 
sterility does not strictly follow systematic affinity. On the 
other hand, how forcibly do they urge, as Mr. Darwin has well 
remarked on certain analogous cases, “that the capacity of two 
species to cross is often connected with constitutional differences 
imperceptible by us, and confined to the reproductive system.” * 
We have seen that dimorphism, as applied to the structure of 
the reproductive organs, is a very general, though not, as has 
been asserted, a universal characteristic of the genus Primula, 
but that several of the species presenting structurally no such 
relations have, on the other hand, their stamens and pistils of an 
equal length. Seeing then that the dimorphic structure, in the 
case of the Primulas, is so invariably correlated with distinctive 
physiological characteristics, we are naturally led to consider the 
nature of the reproductive powers of those species which are 
structurally non-dimorphie, 4. e. those which have stamens and 
pistils of an equal length, and see whether or not these structural 
dissimilarities are connected with any alteration in the func- 
tional characteristics of the species. With the view then of 
illustrating the latter point, I will give the results of a few 
experiments on three of the latter (non-dimorphic) species. 
1. Primula Scotica.—The length of the stamens and pistils of 
this non-dimorphie species varies slightly, as I have already 
stated, under cultivation: there being a regular correlation, how- 
ever, observed, in this variation of the sexual organs, the non- 
dimorphic structure remains unaffected,—the stigma, in every 
case which has come under my observation, being closely ap- 
pressed by the surrounding anthers. From this intimate relation 
of anthers and stigmas, we are naturally inclined to suppose 
that, after the dehiscence of the anthers, the stigmas will be 
liberally supplied with pollen. This is not strictly true; the 
cohesive nature of the pollen-grains still retains them within 
the open lobes, so that if the flowers be carefully guarded from 
external disturbance, the apices of the stigmas, even in their last 
stages of decay, are generally found destitute of pollen-grains. 
* ‘Origin of Species,’ 3rd edit., p. 280. 
