— 227 — 



II. Castration Experiments. 



In the summer of 1903 Mr. Raunkiær and I castrated about 

 20 species of Hieracium. The experiments were carried out in 

 June — July, but as we began so late, we could get only few species 

 of the subgenus Pilosella, because they flower a little earlier than 

 the species of the other common subgenus, Archier actum. 



In the following summers (1904 and 1905) I have continued 

 and extended the experiments. Some few species more of Arcliie- 

 racium have been examined, further several forms of subgen. Pilosella 

 and finally two species of the subgenus Stenotheca, of which all the 

 species (one excepted) are American. I should like very much to 

 get seeds of other species, especially if they belong to more re- 

 markable sections. It is very difficult to obtain species other than 

 the usual ones of Archieracium and Pilosella, which every year 

 appear in the seed-catalogues of the botanical gardens 1 ). I should 

 be glad, if botanists in America would supply me with .seeds of 

 Hieracium, 



The experiments were done in the same way as Raunkiær 

 has described in the case of Taraxacum: We chose flower-heads 

 which are very near to open, and cut off with a razor the upper 

 half of the heads. The part cut off consists of the upper half of 

 the bracts, the coloured part of the corolla, the upper part of the 

 staminal filaments, the anthers and the stigmas together with the 

 upper part of the styles. The surface of the remaining part of 

 the heads is soon covered, more os less, with coagulated latex and 

 shrivel. It is then unreasonable to suppose that an accidentally 

 present pollen-grain should be able to germinate and penetrate 

 through the wounded style to the ovary. Further, supposing that 

 it may have happened sometimes, this explanation fails when we 

 get nearly all the fruits in a head developed. The two only possi- 



embryosacs formed in this way give rise to embryosacs which conse- 

 quently are quite normal and may be capable of fertilisation (compare H. 

 excellens x aurantiacum). But commonly the mothercell of embryosac is 

 displaced and obliterated, while an aposporical embryosac supersedes. 

 The third, but rarer case is, that apogamical embryosacs (as in Taraxa- 

 cum) are developed. In the two last cases the eggcells have the un- 

 reduced number of chromosomes and are capable to form seeds without 

 fertilisation". 



*) Still I have got more than 70 samples of seeds, which now are growing 

 in our Botanical Garden ; but they belong nearly all to the Archieracia and 

 the Piloselloidea. 



15* 



