292 The Pedigree Fauiilies. 



possible to photograph them. This can only be done 

 when few seeds are sown, or when few germinate. 



Fig. 51 shows an example of such a case. The seeds 

 sown were the result of a cross. O. Lamarckiana was 

 crossed with O. nanel.a in August 1899; about 250 of the 

 seeds germinated and about 30 % of the plants they gave 

 rise to were 0. nanella. It is easy to distinguish, in the 

 figure, between the loose rosettes of the parent species 

 and the rosettes of the dwarf by the fact that in the 

 latter the central leaves are more closely packed. Right 

 in the middle there stands rather alone one O. lata easily 

 recognizable by its round (as opposed to pointed) leaves. 

 It was rather covered by its neighbors, so before photo- 

 graphing it I put its leaves over those which were cov- 

 ering it. Otherwise nothing in the group was interfered 

 with. 



The two parents crossed were of pure origin ; the 

 Lamarckiana came from the main line of descent of my 

 experiment (after eight generations of cultivation) the 

 nanella arose from this Lamarckiana race in 1895 and 

 had since bred true for five generations. The lata in 

 Fig. 51, therefore, had no ancestors of a like character 

 at any rate during this period (1886-1899). In the 

 lateral branches of this pedigree however, this form has 

 arisen almost every year. 



The characters of the young plants can be most clearly 

 apprehended for the purposes of identification, from the 

 figures of the rosettes which wiM be given later. Mean- 

 while I propose to describe here the typical form of the 

 adult leaves of the rosettes of the various forms (Figs. 

 52 and 53). The leaves are photographed one-half the 

 natural size and were taken from the plants in the 

 wooden boxes at the beginning of June. The cultures 



