- 165 — 



has become apogamic only at a later period: in almost all ovules 

 typical embryo sac tetrads are formed, and besides in several 

 ovules a vegetative cell becomes the embryo sac. Only in very 

 rare cases the typical embryo sac mother cell is divided with the 

 unreduced number of chromosomes. This makes it evident that 

 a reduction can proceed in most ovules, that is to say that the 

 affinity of corresponding parent chromosomes is so great that they 

 could unite in the synapsis stage. In H. excellens, however, the 

 reduction process is irregular, as only a part of the chromosomes 

 is bivalent in the first division, or in other words, the union of 

 corresponding parent chromosomes is possible only in a part 

 of them. 



Taraxacum shows the next step where the union of chromo- 

 somes in synapsis cannot at all take place, and the division of the 

 embryo sac mother cell is of pure vegetative nature. Perhaps this 

 division corresponds to the second in the reduction division, and 

 this case can then be considered to be a sort of atavismus. 



From this stage it is not far to the embryo sac formation in 

 Antennaria alpina (Juel 5), in which the EMC becomes directly 

 ES with the unreduced number of chromosomes. This is therefore 

 a type adapted only to apogamy. 



The question is now, if the apospory which composes the 

 ordinary manner for the formation of embryo in Hieracium, stands 

 isolated among the higher plants. As I have already mentioned, I 

 contend, that apospory is a more primitive form of apogamy, and 

 that Antennaria alpina represents a much higher grade of adapta- 

 tion to apogamy. If such is the case, — and much leads me to 

 think so, especially the fact that in Hieracium a tetrad division takes 

 place in nearly all ovules — then one must expect to tind apospory 

 in other apogamic forms. In carefully studying the literature upon 

 this subject. I believe I have found cases described, which could 

 very well be understood as apospory. 



In Aster Chamberlain (2) has described a rather peculiar 

 development of the antipodal cells, and a great many of his figures 

 seem to me very extraordinary, especially fig. 1 and 3 of PI. 15. 

 He states, p. 209, that „the antipodal cells are not all alike, the 

 lower one sometimes differing decidedly from the others. It is 

 often much larger than the rest — and its nuclei resemble the 

 endosperm-nucleus rather than the nuclei of the other antipodal 



