464 BREAKDOWN OF GENETIC SYSTEMS 



gives embryo-sacs containing the full set of unpaired chromosomes. 

 This property and the perfection of the mechanism of division 

 in the pollen mother-cells must have been achieved by adaptation 

 in the course of sexual reproduction. Tackholm's objection to 

 this is also no longer valid. He considered that it would lead 

 to frequent aneuploid forms in nature, but we now know that 

 such forms would be largely eliminated {cf. Ch. VIII). The 

 variation necessary for such adaptation will arise from the 

 occasional pairing and crossing over of the chromosomes found by 

 Erlanson (1933) in non-pairing chromosome sets. The present 

 account, therefore, seems to provide a sufficient explanation of the 

 origin and behaviour of the system. 



5. THE ORIGIN OF APOMIXIS IN EXPERIMENT 



The foregoing account has shown the characteristics of nuclear 

 behaviour in apomixis. Apart from these immediate conditions 

 certain events external to the nucleus can be related to the 

 occurrence of apomixis as prior conditions. First we may note 

 that the variations in conditions which are responsible for the 

 change from sexual to non-sexual reproduction may be imposed 

 artificially on organisms with facultative parthenogenesis {cf. 

 Nabours, 191 9 on Paratettix and Apotettix). 



(i) Pseudogamy. In some species of plants and animals the male 

 gamete functions in exciting the development of the unfertilised 

 and unreduced ^gg in diploid parthenogenesis or of nucellar embryos 

 in apospory. This is known as pseudogamy. We have already 

 seen how non-recurrent parthenogenesis is stimulated in plants in 

 this way. 



In a nematode, Rhahditis "XX," Belar (1923) found that the 

 sperm, although irregularly formed (in part of the ovarian tissue), 

 entered the egg and stimulated its development, probably by the 

 contribution of its centrosome — the middle-piece of the sperm — 

 although no fusion of the nuclei occurred. 



An analogous process is found in various monocotyledons, e.g., 

 Zephyranthes texana and Allium odorum (Modilewski, 1930). Here 

 the pollen is normally developed and one generative nucleus fuses 



