246 Ostcnfeld. 



plant concerned has wholly lost the power of developing 

 fertilized seeds or some flowers of each head require fer- 

 tilization in order to produce seeds, while others develop 

 seeds apogamically. I think that this way of solving the problem 

 is better than that of the microscopical examination. Then, if only 

 a small number of the flowers require fertilization, it is very difficult 

 by means of microscopical investigation to find the rare stages in 

 the development of the embryo-sac, which show the phases of chro- 

 mosome-reduction and which consequently prove the necessity of 

 fertihzationi). 



A difficult thing to find out is, how great a numerical material is 

 required to get trustworthy results from these investigations, and here 

 I am sorry to say that my material is hardly sufficiently extensive. 

 My tables (I — III) show just that the greater the numbers are, the 

 better the percentages agree. 



In igo6 I could enumerate 14 species of Archieradnm in which 

 I had found apogamy. As the tables (I and II) show, the number 

 of forms has now increased to 60. The seeds of most of these I have 

 had from various Botanical Gardens from their seed catalogues. The. 

 names under which these seeds were received have in many cases been 

 wrong, and to get better determinations I applied to our best hie- 

 raciologist, Dr. H. Dahlstedt of Stockholm and asked him to deter- 

 mine my experiment-plants. I take here the opportunity of expressing 

 my sincerest thanks to him for his obliging answer to my request. 

 His determinations have led to the number of examined forms being 

 considerably diminished; it comprehends only about 33 species, taken 

 in a very wide sense. 



The species examined belong to very different groups of the 

 subgenus, and considering that among all the experiment plants only 

 one species, viz. //. virga aurea Coss. — setting aside the forms 

 of H. umbellaluni L. and the nearly allied H. lactaris Bert. — has 

 not been found to be apogamic, I think I am justified in saying 

 that almost all the numerous species of the subgenus 

 Archieracium are apogamic. There is no reason to go through 

 the list of the single species, it must be sufficient to refer to the 



1) I think it hardly correct, therefore, when S. Murbeck (1904, p. 294) concludes 

 that, because in three Hieracium species examined he never found pollen tubes in the 

 style nor in the micropyle, "die drei betreffenden Arten sehr wahrscheinlich stets 

 parthcnogenetisch sind". The result is, in all probability, right, but the conclusion is 

 not allowable. 



