MINUTE STRUCTURE OF PLANT HYBRIDS. 269 



more truly vegetative regions that that strikingly diversified intermixture of tissues is 

 found, which causes it to differ from all seed hybrids that I have studied. 



But the parents that were used for the production of the graft hybrid in the present 

 instance are very unlike in habit, colour, &c, so that even a seed hybrid from them would 

 be more than worthy to rank alongside Philageria ; and if the latter is sterile, need we 

 wonder that this graft hybrid is also ? Were two species chosen, however, of close 

 affinity, we suspect that a graft fusion as intimate as that just examined might bear 

 perfect fruit in the hybrid part, and that such might reproduce a progeny as perfectly as 

 do some of the seed hybrids. Thus a graft cross of TJlmus campestris and U. montana, 

 of Fagus sylvatica and F. ferruginea, or of Betula alba and B. nigra, would probably 

 give the desired result. This view is greatly strengthened when we recall the fact, 

 already alluded to, that nearly all the pollen cells of C. Adami are good in appearance, 

 though the ovules, according to Professor Caspary, are mostly monstrous. 



The flowers on the parent branches of the composite organism normally bear fruit 

 almost or quite as abundantly as if each had grown independently, and the seeds give 

 rise to plants like the parents, but the flowers on the hybrid branches never produce ripe 

 fruit. This, we think, is a strong argument in favour of the hypothesis advanced in the 

 latter part of this paper to account for relative sterility, if in the present instance we 

 further suppose that the original plant resulted as a graft shoot from accidental union of 

 the halves of a bud of each parent. Thus, if the grafter, in preparing the stock and 

 shield, cut in half a vegetative bud along the margin of each where future union was to 

 be effected, not only would the shield graft produce pure shoots from pure buds over its 

 surface, but if union of the cellular tissue of each half bud of stock and graft respectively 

 was accomplished, the product would be a composite bud, one side of which would ulti- 

 mately form branches of C. Laburnum, the other of C. purpureus ; but along the 

 junction surfaces union of protoplasm, of nuclear threads, and of chromatic substance 

 might be effected so intimately that a hybrid tissue growth would ensue, showing admix- 

 ture of structures characteristic of both parents. 



We are compelled to assume that a union of nuclei has taken place in view of the 

 important role played by the nucleus in cell life, and also by the close resemblance which 

 the flowers of C. Adami have to those of a seed hybrid which have thus resulted. 



Now, if such a composite growth were isolated and propagated, as was actually done 

 according to M. Adam's account, the more copious and complex the branching and new 

 bud formation became the more perfectly would admixture of the segregated cells of 

 yellow, purple, and red parts become, without the necessity of their losing individuality. 

 But just such an intergrowth and admixture would explain the histological peculiarities 

 which we have met with in the vegetative parts of C. Adami. Though actual experiment 

 and observation alone will decide the point, it seems to me essential for the production 

 of such graft hybrids that halves of two buds should be united. Some have supposed 

 that the formation of adventitious buds from the cambial layer in the region of graft 

 union would best explain the requirements of the case. In view of observations such as 



