540 GENERAL REVIEW. 



flowered group, while D. tatula, querrifolia, and some others, 

 bear purple flowers. D. l<zvis and D. ferox will hybridise 

 reciprocally, and in 1863 M. Naudin raised 60 individuals of 

 D. Icevi-ferox, and 70 of D. feroci-lcevis. The offspring of this 

 reciprocal cross attained "the most complete development, 

 and were so perfectly like each other that the two sets might 

 easily have been regarded as one." " This is a new confirma- 

 tion of what I have already announced (see ' Comptes Rendus 

 de 1'Acad. des Sciences,' 1862), that there is not a sensible dif- 

 ference between reciprocal hybrids of two species, and that in 

 the first generation the hybrids of the same origin resemble 

 each other as much as individuals of pure species from the 

 same sowing." " But these 130 hybrids presented a fact which 

 was quite new to me : if they perfectly resembled each other, 

 they differed strangely from the two species from which they 

 were derived. They had neither the stature, the habit, the 

 flowers, nor the fruit of their parents ; there was not even any- 

 thing intermediate between their forms, which were so well 

 known and so decided. Any one who did not know the origin 

 of these hybrids would not have hesitated to make a new 

 species of them, and, what is worth notice, would have classed 

 them in the violet series, for all had the flowers of this colour 

 and brown stems. Notwithstanding, as I said above, the two 

 parent species belong to the group characterised by green stems 

 and white flowers. 



" In the face of this unexpected result, one might have been 

 tempted to believe that two species, intermarrying, might im- 

 part to their progeny characters which they do not possess 

 themselves ; but such a conclusion was too paradoxical to be 

 accepted without a re-examination. I resolved, therefore, to 

 recommence the experiment the following year, observing at 

 the same time more closely, not only the hybrids, but also the 

 species from whence they were derived. In 1864 I again 

 sowed D. Icevi-ferox and feroci-lavis, and by their side the par- 

 ents, D. ferox and D. Icevis, in a state of purity. 36 new 

 plants of D. l&vi-ferox, and 39 of D. feroti-l&vis, reproduced 

 all the identical features of their brethren of the preceding 

 year. They had the same brown stems, the violet flowers, and 

 thorny fruit. But what I had not previously remarked in D. 

 ferox of a pure strain, the stem at the moment of germination 

 is of a deep purplish violet. This vivid tint extends from the 

 root to the cotyledons, where it suddenly ceases, giving place 

 to the clear green tint ; but it remains during the whole exist- 

 ence of the plant on the part which it occupies, and where it 

 traces a coloured circle. From this moment all was clear : it 



