APOMIXIS 



317 



Three cases of androgenic haploids, in which the male nucleus 

 alone is concerned in the development, are also on record, but no 

 developmental studies have been made and the evidence is entirely 

 genetical. Kostoff (1929) pollinated Nicotiana tabacum var. macro- 

 phylla, having 72 chromosomes, with N. langsdorffii, having 18 

 chromosomes. Out of about 1000 seedlings resulting frcm this 



Fig. 180. Haploid parthenogenesis in Epipactis latifolia (s = synergid; o = egg; 

 t = pollen tube; p = secondary nucleus). A, B, embryo sacs showing division of 

 haploid egg; pollen tube t is still undischarged. C, embryo sac has not received 

 any pollen tube, but egg and secondary nucleus are ready to divide. D,E, meta- 

 phases in first division of haploid egg, showing 20 chromosomes. (After Hagerup, 

 1945.) 



cross, one reached maturity. This showed none of the characters of 

 N. tabacum but strongly resembled a dwarf N. langsdorffii. The 

 number of chromosomes turned out to be 9, strongly suggesting its 

 origin from a male gamete of N. langsdorffii. 



In the same year Clausen and Lammerts (1929) reported a parallel 

 case in N. tabacum. They crossed N. digluta 9 , an allohexaploid 

 having 72 chromosomes, with N. tabacum <?, having 48 chromosomes, 

 and obtained a plant with 24 chromosomes. This plant agreed 

 morphologically and cytologically with other haploid tabacum plants 



