372 METAMORPHOSIS 



sexual-cells also which arise from stamens and carpels of one and the same 

 flower can scarcely possess any very marked individual characteristics in their 

 initials. Still, it is known that there are numberless arrangements to be observed 

 in flowers by means of which ' self-pollination ', that is to say, transference of 

 pollen to the stigma of the same flower, is prevented, and by which cross- 

 pollination between neighbouring branches or even between neighbouring 

 plants is facilitated (Cross-fertilization, DARWIN, 1876). Moreover the gametes 

 formed in one gametangium in certain Algae are said not to fuse with each 

 other (SxRASBURGER, 1900 b, 306). 



If we now try to follow up more closely the complete combination of initials 

 in fertilization we at once come face to face with a great difficulty, for system- 

 atic experiments are wanting in which well-chosen examples with marked 

 individual characteristics have been crossed and the offspring thoroughly 

 studied. To carry out such researches would be very difficult, since individual 

 differences in plants are not, as a general rule, very well marked. For this 

 reason we must go further afield and study hybrids between different plant 

 relations (races, varieties, species) which show more easily recognizable differ- 

 ences between each other. As to such crosses there exists a voluminous litera- 

 ture. The existence of such races we take for granted the next lecture will 

 deal with their mode of origin in the present lecture we will deal only with the 

 results of crossing individuals belonging to two different races that is to say, 

 with the production of hybrids (the older literature relating to these will be 

 found in FOCKE, 1881). [For more recent literature see DE VRIES, 1903.] 



So far as we are aware, FAIRCHILD, in England, in the year 1717, was the 

 first gardener to raise a hybrid, inasmuch as he pollinated the stigma of Dianthus 

 caryophyllus with pollen for D. barbatus. Among botanists KOLREUTER (1761) 

 was the first to conduct experiments in hybridization for years on a large scale. 

 He was interested in hybrids from the point of view of their bearing on the 

 sexuality of plants, a subject which had been much disputed. His first hybrid, 

 Nicotiana rustica $xN. paniculata a*, flowered in the summer of 1761. Since 

 then innumerable hybrids have been produced for scientific and horticultural 

 purposes ; many of them have arisen naturally, and yet recent developments 

 in our science indicate that we have only just crossed the borders of the sub- 

 ject, and that in this province a wide and interesting field of research lies open 

 to us (compare DE VRIES, 1900 ; CORRENS, 1900 onwards ; H. TSCHERMAK, 

 1900 ; summarized by CORRENS, 1901 a and 1903, [1905]). 



It is not possible to produce hybrids from any two plants selected at 

 random, because the capacity for forming hybrids is generally restricted to 

 nearly-allied plants. Rarely do we find species of different genera capable of 

 hybridizing ; less frequently still are hybrids themselves capable of crossing. The 

 more closely related the plants, the easier it is, as a rule, to produce a hybrid. 

 Still, the capacity for hybrid-production does not, in any sense, run parallel with 

 systematic relationship. It is very noteworthy that in certain cases the 

 hybrid A $ x B a* may occur while the reciprocal B $ x A c/ 7 is impossible. Thus 

 Mirabilis jalapay is easily crossed with M.longi flora a* 9 whilst it is impossible to 

 fertilize M . longiflora <j> by M . jalapa a*. Such a fact at first sight seems almost 

 incomprehensible, but it is suggested that the success of the cross depends not 

 only on the capacity of the sexual-cells to fuse with each other but on the possi- 

 bility of the approximation of these cells. It is usually assumed that the 

 pollen-tube of M . jalapa is too short to grow through the much longer style of 

 M. longiflora. This may be true, but there is also another possibility, recently 

 drawn attention to by BURCK (1900). BURCK found that many stigmas contain 

 substances which are capable of stimulating the pollen-grains only of some, but 

 not of all, species of the same genus to develop. It must be assumed that the 

 failure of many experiments in crossing is due only to the fact that the pollen- 

 grains of the male parent are incapable of germinating on the stigma of the 



