172 



F. HUSSEIN 



chromosomes measured about 1 ft. Constrictions, seldom visible, were mainly sub- 

 terminal or median, with a few submedian ones. There was, however, a longer pair 

 measuring 2 /a which seemed always to have a median constriction. In meiosis of the 

 single flowers, a large diamond-shape bivalent with a median attachment was always 

 clear and would correspond with the longest pair of somatic chromosomes (Fig. 1 b). 

 Multivalents were found in about 2*5 per cent of cases and were either tri- or quadrivalents. 



The results of anaphase separation were nearly always equal. These results agree 

 with Lawrence's (1931) description, from which he concluded that the species is an 

 allotetraploid. 



Fig. 1. (a) Somatic chromosomes of 2n = 30 of a single, double and " Arabis " type; (b) Meiotic chromo- 

 somes of a single 2n = 30. 



Plants with 2n — 56 had either single or semidouble flowers. The study of somatic 

 and meiotic plates of the two types {see Fig. 2 a and b) showed no major differences. 



The differences in size and shape among the chromosomes of the same complement 

 were also few, and may be correlated with those found in the 2n = 30 plants. However, 

 owing to the smallness of the chromosomes, it was not possible to identify them individually 

 and thus one could not judge the amount of similarity between the two races. The result 

 of meiotic anaphase separation in 2n = 56 was also equal. This race may be considered 

 an aneuploid of an octoploid. Whether it is an auto- or allopolyploid of the 2n = 30 

 race one cannot decide unless the meiotic behaviour of a hybrid and of an artificially 

 produced 2n = 60 is examined. Except for the aneuploidy, the case is similar to that 



