120 PLANT PHYSIOLOGY 



of the individual, it is a matter of chance whether the paternal or the maternal 

 initials make themselves evident. One cannot assume, however, that only 

 one initial is present in a single twig on the contrary, all cells of such a mosaic 

 hybrid must carry both paternal and maternal initials, since, according to 

 the chromosome theory, segregation can make its appearance only in reduc- 

 tion division, i. e. when the sexual organs are formed. There is a plant which 

 is doubtless a hybrid between Cytisus Laburnum and C. purpureus, and which 

 is essentially intermediate between the two parent species. In this hybrid 

 Cytisus Adami we find, extremely frequently, branches which have taken on 

 completely the characters of C. Laburnum, more rarely those of C. purpureus 

 (comp. Figs, in the Bonn Text -book). These reversions are really segrega- 

 tions ; the Laburnum twigs are perfectly fertile (while C. Adami is sterile), and 

 from their seeds the pure Laburnum-type arises, without any suggestions either 

 of Adami or purpureus. Now the reversion in question cannot very well be 

 associated with a reduction division, and hence segregations must obviously 

 take place independently of such divisions. We must go still further and say 

 that the reversions in this case have no connexion with either nuclear or cell 

 division. Occasionally one notices that only one half of a bud has reverted ; 

 and then the branch arising from it has on one side the characters of Laburnum, 

 on the other of Adami; further, the boundary between these may be in the 

 middle of a leaf or a flower (BRAUN, 1851) ; in such cases it is perfectly obvious 

 that the segregations are not limited to derivatives of one cell (BEIJERINCK, 

 1901). Owing to this fact it would appear certain that the distribution of the 

 initials in the cells is not dependent on nuclear division. 



The origin of Cytisus Adami is obscure. Since its first description by 

 ADAM it has been found impossible to produce the hybrid anew. We do not 

 know whether it is an ordinary hybrid or a so-called graft-hybrid, i.e. a hybrid 

 arising from a vegetative union of two parents. As it is not possible at all to 

 produce such graft-hybrids (VoCHTiNG, 1892), statements as to their exis- 

 tence have been sceptically received. KOHNE (1902), however, has drawn 

 attention to a hybrid produced at Bronvaux between Crataegus and Mespilus, 

 which arose close to the point of concrescence of a Mespilus scion on a Cra- 

 taegus stock. According to NOLL (1905) it is scarcely possible to doubt that in 

 this case we are dealing with a genuine graft-hybrid. NOLL assumes that the 

 bud from which the hybrid was formed might be referred back to a fusion of 

 two vegetative cells which had accidentally entered the callus. According 

 to MIEHE (1901), KORNICKE (1901), and NEMEC (1904), a transference of 

 nuclei from one cell to another in the neighbourhood has been observed to 

 take place in other cases, and that this occurs frequently after certain stimuli 

 have been applied. But without such a stimulus it has been shown to take 

 place in certain apogamous fern prothalli (FARMER and DIGBY, 1907), and in 

 that case, after the transference, a fusion of the nuclei and the formation of 

 a fem-sporophyte followed. Hence we cannot deny that in a callus of two 

 graft-symbionts such a fusion may occasionally take place leading to the 

 formation of a vegetative hybrid. But because the chances are so greatly 

 against such fusions, graft-hybrids must be correspondingly rare. 



The Bronvaux hybrids (comp. NOLL, 1905) are distinguished in two 

 points from ordinary hybrids produced by crossing. In the first place, they 

 showed segregations similar to those of Cytisus Adami, and secondly, the three 

 shoots formed on the callus differed among themselves, although they were all 

 intermediate in character between their parents. One of them (forma Jouini, 

 NOLL) is of especial interest because it had at first quite the characters of 

 Crataegus, and then gradually assumed those of the hybrid. This shows that 

 the initials of a character can appear in cells after these are adult, and hence 

 must be independent of the nucleus. One might indeed say that protoplasmic 



