OF DTMOKPIIIC AND TRIMOKPIIIC PLANTS. 435 



has once appeared, it is inlierited witli remarkable force. Plants 

 which have become equal-styled, and have thus lost their dimor- 

 phic structure, are perfectly self-fertile, being quite as fertile as 

 ordinary plants when legitimately crossed. This being the case, 

 and as the variation so often arises, it may be asked why has it 

 not occurred under nature and been naturally selected or pre- 

 served. The answer, no doubt, is that such plants would be emi- 

 nently liable to long-continued self-fertilization, which certainly 

 entails a weak constitution *. 



As the great majority of plants of all kinds and even some 

 species of Primula t are non-dimorphic, the loss of dimorphism in 

 the equal-styled varieties may be attributed, as Mr. Scott has re- 

 marked, to reversion to the primordial condition of the plant ; and 

 this explains the force with which this modification is inherited. 

 We have also seen in illegitimate plants descended from the long- 

 styled P. sinensis that which appears to be another case of reversion, 

 namely, the small size and wild aspect of their flowers. Now I 



have elsewhere :|; given abundant evidence showing that the off- 

 spring of crossed species and varieties are eminently liable to re- 

 version. Hence in the cases in which illegitimate birth appears 

 to have been the exciting cause of reversion, illegitimacy has acted 

 like hybridization. The parallelism in this particular instance is 

 close ; in a future paper I shall show that the common Opclip is a 

 hybrid between P. veris and vulgaris] and I have seen short-styled 

 wild Oxlips which had become strictly equal-styled, and others 

 which exhibited gradations in the length of the pistil, but not in 

 the roughness of the stigma, leading to this same state, like the 

 gradations described under P. sinensis and veris. 



Although there may be some doubt with respect to the paral- 

 lelism between illegitimate unions with their illegitimate oflfspring 

 and hybrid unions with their hybrid offspring, in regard to the last 



two subjects discussed, namely, the disturbed proportions of the 

 sexual forms and sexes, and the appearance through reversion of 



equal-styled varieties, there can be no doubt that the parallelism 

 is so close as to amount almost to identity in the following chief 

 characteristic points, namely : — the various grades of lessened 

 fertility up to complete barrenness — the fertility innately differ- 



* See my work on the * Yariation of Animals and Plants under Domestica- 

 tion/ 18(>8, vol. ii. chap, xvii., and especiaUy p. 128. 



t Mr. J. Scott, ** on the Reproductive Organs in the Primulaccae/* Proc. Linn. 

 Soe. Bot. vol. viii. (1864) p. 78. 



* Variation of Animals and Plants under Douicstication, vol. ii. chap. xiii. 



