222 



HARDWICKE'S SCIENCE-GOSSIP. 



[Oct. 1, 1S70. 



merits show that in order to insure complete ferti- 

 lization the two forms must be intercrossed, pollen 

 from the anthers of the long-styled form being ap- 

 plied to the stigma in the short-styled form, and 

 vice versa : thus out of one hundred flowers of 

 cowslips fertilized by pollen from the same form, 

 the average weight of seed proved to be eleven 

 grains ; whereas one hundred flowers fertilized 

 with pollen from the other form produced thirty- 

 nine grains weight of seed, or more than three times 

 as much. 



Further experiments on the fertility and per- 

 manency of the offspring of these different unions 

 reveal the remarkable fact that the offspring of the 

 union of two plants of the same form are, to all 

 intents and purposes, hybrids formed within the 

 limits of one and the same species ; so that in the 

 species of Primula, although the flowers are struc- 

 turally hermaphrodite, or presenting both sexes, 

 they are in reality dioecious or unisexual. 



That the species may eventually become com- 

 pletely dioecious, and that this dimorphic condition 

 in the sexual organs maybe a step towards this end, 

 does not seem improbable. Many botanists consider 

 that there is a tendency in all plants towards a sepa- 

 ration of the sexes; and the fact that several plants, 

 normally dioecious, present occasionally bisexual 

 flowers, and more frequently flowers in which the 

 pistils or stamens, instead of beiug entirely sup- 

 pressed, are merely aborted or rendered rudimentary, 

 seems to me to afford a strong argument in favour 

 of such a view ; and I do not see much difficulty in 

 the belief that the stamens in one form, and the 

 pistil in the other form of the different species 

 belonging to the Primrose family, will eventually 

 become suppressed, and the sexes completely 

 separated. 



But this, and the no less interesting question 

 as to the form or forms from which our prim- 

 roses and cowslips have sprung, are rather beyond 

 the scope of this paper. It is evident that what- 

 ever advantage there is in a separation of the 

 sexes, and intercrossing of individuals, is, at least, 

 partially gained by this dimorphism and cross -fer- 

 tilization of the two forms. 



Dimorphism is by no means confined to plants 

 of the Primrose family. Several species of Linum 

 also present two forms. In Linum grandijlorum the 

 difference is confined to the pistils, which are shorter 

 and more divergent in one form than in the other. 

 In Linum perenne the long-styled plants have stamens 

 only about half the length of those in the short- 

 styled forms. 



In this case, as in that of the Primrose, the 

 forms must be crossed in order to insure fertility, 

 the pollen-grains of both forms being as unable to 

 effect fertilization when applied to the stigma of 

 their own form, as if they had come from the 

 anthers of a totally distinct species. 



Several species of Lythrum are also dimorphic. 

 Of the two species indigenous in England, Lythrum 

 hjjssopifolium is stated by Vaucher to be dimorphic ; 

 Darwin, however, is of the contrary opinion : the 

 plant is not a common one, and I have never seen 

 it. The other British species, Lythrum Salicctria, 

 the common purple loosestrife, presents three dis- 

 tinct forms, each differing from the others in the 

 length of the pistil, and each having two sets of 

 stamens differing from each other in appearance 

 and function. The styles in the three forms differ 

 greatly in length. In the long-styled form the style 

 is one-third longer than in the mid-styled, and more 

 than three times as long as in the short-styled form, 

 and corresponds in length to the longest stamens 

 in the other two forms. In like manner the length 

 of the style in the mid-styled form corresponds with 

 the length of the long set of stamens in the long- 

 styled, aud the shorter stamen in the short-styled 

 form, and the style in the short-styled form corre- 

 sponds in length with the short stamens in each of 

 the other two forms. There are also other minor 

 differences in the three forms, in the colour of the 

 filaments and pollen, and number and size of the 

 seeds. In the long-styled form, the filaments and 

 pollen are yellow ; in the other two forms the long 

 stamens have their filaments pink, and pollen green, 

 the short ones being yellow. In all the forms the 

 size of the pollen-grains from the long stamens 

 slightly exceeds that from the short ones. 



As in the dimorphic plants I have before men- 

 tioned, so in this trimorphic Lythrum very import- 

 ant and remarkable functional differences are added 

 to the differences in structure, and it has been found 

 by careful experiments, that for the full fertilization 

 of each form it is necessary that the pistil should be 

 acted upon by pollen from the stamens of correspond- 

 ing length in one of the other two forms — in other 

 words, the longest style is only fertilized by pollen 

 from the longest stamens of one of the other two 

 forms, the middle style from the middle stamens, and 

 the short style from the short stamens. In all three 

 forms the pistil is only very feebly, or not at all, acted 

 upon by pollen from the same form ; the greater 

 the inequality in length between the pistil and the 

 stamens the more completely sterile being the union 

 between them. 



Eor the fertilization of all the flowers I have 

 mentioned, insect-agency is more or less absolutely 

 necessary, and all are frequently visited by bees, 

 moths, and other insects. It is easy to see that, if 

 a bee visits, for instance, the flowers of the purple 

 loosestrife, the longest stamens rub against the 

 abdomen and hind legs, as does the stigma in the 

 long-styled forms, aud so in all the forms the stigma 

 comes into contact with just that part of the insect's 

 body which is sure to be covered with the right 

 kind of pollen for its fertilization. 



No contrivance could be better adapted to insure 



