On the Genetics of Tetraploid Plants in P. sinensis. 485 



accompanied by others at first sight entirely paradoxical (as, for example, 

 the fact that the ostensible recessive may throw the dominant), and the 

 whole series may be regarded as of special significance in view of the 

 association with the doubled condition of the cell-constituents. Moreover, 

 in the tetraploid Primulas, the reduplication affects not merely the factors 

 for isolated characters, but extends simultaneously to the factors for all 

 the characters so far investigated. 



The tetraploid giants with which I have worked are of two distinct races. 

 One of these, which will be referred to as the GX race, consists of the 

 progeny of a plant kindly given to me by Messrs. Sutton and Sons. The 

 other race (G-T race) arose in the course of my own experiments. Two 

 non-giant diploid plants were crossed together reciprocally. The Fi from 

 one of these crosses gave a perfectly normal F 2 , consisting of non-giant 

 plants among which all the expected classes of offspring were represented 

 in numbers closely approximating to expectation. The Fi from the 

 reciprocal cross gave no seeds in a cross with one of its parent races and 

 gave only four plants as a result of self-fertilisation. These four plants 

 were giants, and from one of them the GT race has been bred. 



Up to the present time, neither the GX nor the GT races of giants have 

 given any fertile seeds in crosses with various non-giant (diploid) races, 

 whichever way the -crosses were made. In this respect they differ from 

 a diploid giant race, with which I have worked, which proved quite 

 fertile with non-giants. It was this difference in behaviour which led to 

 the discovery of the tetraploid nature of the GX and GT races. 



In the tetraploid plants the chromosomes are naturally more crowded on 

 the spindles than they are in diploid plants, but in polar views of the 

 spindles of either of the maturation divisions there is no difficulty in 

 determining that the number of chromosomes is normally 24 (compared 

 with the 12 chromosomes found in diploid plants). In the somatic 

 mitoses the chromosomes are longer than those of the maturation divisions 

 and exact counts are difficult to make, but a number of counts have given 

 numbers approximating to 48. The maturation divisions sometimes show 

 some degree of irregularity, one or two chromosomes lagging behind the 

 others in the movement to the poles, but I am not yet able to say whether 

 fertile germ cells having more, or fewer, chromosomes than 24 are ever 

 formed. There is no visible difference between the chromosome groups of 

 the thrum-eyed (short-styled) and pin-eyed (long-styled) plants. 



The two plants which were the progenitors respectively of the GX and 

 GT races each possessed its own series of dominant characters, in respect 

 of which its origin would indicate that it was heterozygous. In the 



