- 164 - 



consequently necessary for embryo formation, which has also been 

 proved by Ostenfeld in his castration experiments. In some 

 other species, as for instance in H. excellens, there are also deve- 

 loped quite typical ES with a reduced number of chromosomes, 

 which probably only develop embryo after fertilization. We are 

 justified in assuming that the hybrid seeds in Ostenfeld' s experi- 

 ments originated from flowers with typical ES. 



However the greater part of Hieracinm are apogamic accor- 

 ding to his investigations, that is to say, that they can develop 

 seeds without fertilization. 



In the apogamic species of Hieracium I have found, but very 

 seldom, an ES-formation similar to the one described .by Juel in 

 Taraxacum : the EMC was divided only once and with the unreduced 

 number of chromosomes. 



The greater part of the embryo sacs in the ovules of Hiera- 

 cium are, however, of aposporic origin, as a vegetative cell deve- 

 lops into an embryo sac, which consequently has the unreduced 

 number of chromosomes. Two synergides, one egg cell, three 

 antipodes and two polar nuclei are formed. The two last men- 

 tioned, sooner or later, unite and make a central nucleus. 



The aforementioned vegetative cell is situated differently in 

 different species. In H. fiagellare it is often a cell from the inte- 

 gument or from the chalaza-region. In H. excellens it is generally 

 a cell quite near the tetrad, in H. aurantiacum generally an epi- 

 dermis cell of the nucellus. 



Simultaneously with the development of this vegetative cell to 

 an embryo sac, the embryo sac mothercell is, as usual, divided 

 into tetrads with the reduced number of chromosomes, and a 

 normal ES commences to develop. This, however, most usually, 

 is sooner or later destroyed by the encroaching aposporic embryo 

 sac. Still in some cases two embryo sacs could be fully deve- 

 loped: one probably typical and the other aposporic. In the 

 same head some ovules have a typical and others an aposporic 

 embryo sac. 



I think that in this combination in the same head of typical 

 and aposporic embryo sacs we have the explanation why H. ex- 

 cellens, in spite of its being apogamic, can form hybrids. 



A comparison of the hitherto known apogamic Compositae is 

 of certain interest. It seems to me that Hieracium shows the 

 most primitive form of apogamy, or in other words that this genus 



