474 BREAKDOWN OF GENETIC SYSTEMS 



in parthenogenetic plants : the prophase begins after a prolonged 

 resting-stage and the chromosomes are contracted very little 

 more than in mitosis (e.g., in Archieracium, Eupatormm, Antennaria, 

 Bergman, 1934, cf. Gustafsson, 1935 ; Leontodon hispidtts, abnormal 

 plant, Bergman, 1935). 



We see then, on cytological grounds, that the suppression of 

 meiosis in these plants is conditioned in its simplest form by the 

 mechanical relationships of the chromosomes, in its more advanced 

 forms, on the other hand, by a genotypical property. This property 

 acts by removing the characteristic precocity in the prophase and 

 with it the essential characteristic of meiosis, the pairing of chromo- 

 somes. It may have more or less of the adventitious characteristics 

 of meiosis, which are secondary to this pairing and do not hinder the 

 regularity of the substituted mitosis. 



Thus three conditions appear to determine diploid parthenogenesis 

 in its most advanced form in the Eu-hieracia and elsewhere : (i) A 

 genetic property, particularly common in some genera, conditioning 

 parthenogenetic development of the embryo-sac ; (ii) hybridity 

 or autopolyploidy conditioning the original formation of unreduced 

 embryo-sacs, as found in the Pilosella section ; and (iii) a genetic 

 property acquired later by adaptation, perfecting the mechanism 

 by which suppression takes place and probably consisting in a 

 derangement of the ordinary time relationships of meiosis ; the 

 first stage of this is found in Boreale, last stage in Psendoillyricum. 



(ii) Adaptation and the Mechanism of Variation with Partheno- 

 genesis. A difficulty arises when we come to consider how this 

 third condition can be attained by adaptation in a parthenogenetic 

 organism. This difficulty has been felt in another connection. 

 Genera of plants in which apomixis is general, commonly show a 

 remarkable potymorphism. This is particularly true of Rubus and 

 Eu-hierachim. In En-hieracium Ostenfeld found 60 partheno- 

 genetic species and only three sexual ones (cf. Rosenberg, 1917). 

 How has such a remarkable polymorphism arisen in the absence of 

 the two recognised means for the distribution and recombination 

 of variations, viz., meiosis and fertilisation ? 



Ostenfeld and many others have sought to explain the phenomenon 

 as the result of mutations (192 1). As a rule the progeny of a 



