VEGETABLE HYBRIDITY. 87 
pistils of each species were impregnated with the pollen of the other. 
From the species D. Icevis and ferox thus douhly crotsed, the autlior 
obtained in 1863 sixty youni^ plants of i). l^vi-ftrox, and seventy of 
D.feroci-lrevis, The whole of these 130 plants grew freely, and were 
so perfectly similar in appearance that the two sets coidd not Lave 
been distini^iiished, tlie entii'e collection of hybrids beinir as homoo-e- 
neons and uniform as if they had been a group of individuals of a fixed 
species or a pure and distinctly marked race. 
On the other hand, to JM. Naudiu^s sui'prise, these hybrids presented 
no appearance of being intermediate between the two well-marked 
species from which they were derived, so that any one ignorant of their 
origin would not have hesitated to regard them as forming a distinct 
species : and, curioiisly enough, whilst both the parents belonged to 
the section with green stems and white flowers, the hybrids would be 
referred to the other group, their flowers being violet and their stems 
b 
rown. 
This result was so unexpected and paradoxical that M. Naudin re- 
solved to repeat his experiments, and this year he made a new sowing 
both of the hybrids and of the parent species. He obtained thirty-six 
new plants of D. Iceoi-ftrox and thirty-nine of D. feroci-lcevisy w^hich 
were identical wich their predecessors of 1863, having the stems brown, 
the flowers violet, and* the fruit spinose. But the sowing of B, ftrox 
furnished an explanation of this curious f<ict, for the author found 
that, at the moment of germination, the stem is of a deep violet-purple 
tint from the root to the cotyledons, and that this coloration persists in 
its oriaiual place throughout the life of the plant, forming a coloured 
circle round the stem. Thus the tendency to coloration seems to rer 
side in the B, ferox, although here it is reduced to a rudimentary 
state ; in the hybrid it becomes enormously increased, pervading all 
parts of the plant, and especially influencing the flower. 
The second generation presented variations of a different and still 
more remarkable kind. The seeds of the above-mentioned hybrids 
sown last spring furnished nineteen plants of B, feroci-lavu and 
twenty-six of B, Icevi-ferox. But in spite of the great similarity of 
their parents, these plants presented a most astonishing diversity of 
forms, so that out of the forty-five plants composing the two sets, no 
two were exactly alike. They differed greatly in size (some being four 
times as large as others), in general aspect, in form of leaf, in the co- 
