180 Genetic Behaviour of the Hybrid Primula Kewensis 



between size and chromosome number appears definite, but in the 

 diploid giant there is a general increase in size without a doubling of the 

 chromosome number. In Kewensis the tetraploid varies considerably 

 in size, the small plants being no larger than the diploid form ; the 

 chromosomes of one of these small plants were counted by Miss Thomas, 

 and were found to be 86 diploid number, 18 haploid number. It 

 is doubtful whether the tetraploid attains a greater size than well- 

 grown plants of verticillata, certainly not greater than might be 

 accounted for by the increase of vigour often seen in hybrids. It is 

 desirable that further chromo.some counts should be made, meanwhile 

 we must assume that number of chromosomes is not directly associated 

 with the size of plant. 



In the tetraploid P. sinensis, Gregory found that the factorial 

 system is doubled, while in the diploid giant this is not so. We cannot 

 say if the tetraploid Kewensis possesses a doubled factorial system, for 

 we have no diploid form with which to compare it. The only characters 

 which exhibit anything like normal segregatii)n in Keiuensis are the 

 various degrees of mealiness, and the phenomena met with in the 

 inheritance of these characters might be equally well accounted for by 

 the doubling of a few pairs of factors or by the presence of several factors. 



In P. sinensis fertilization cannot be effected between the tetraploid 

 and diploid forms, but in Kewensis this is not the case. 



The genetic behaviour of CE. gigas exhibits certain puzzling 

 features. Dwarf forms are thrown by it in the proportion of 40 per 

 cent., but we are not told if these are diploid or tetraploid. A narrow- 

 leaved form appears but de Vries considers the leaf breadth to be 

 a fluctuating character. Crosses between gigas and other forms have 

 given various results in the hands of different workers. It seems 

 probable that individual plants differ in their genetic constitution, in 

 spite of the fact that they breed true to gigas on self-fertilization. (For 

 summary of facts and discussion see Heribert-Nilsson'.) 



)SUMMARY. 



I. From the cross P. verticillata by floribunda and the reciprocal, 

 plants resembling the female parent are generally obtained. Their 

 origin as the result of parthenogenesis cannot be held to be disproved, 

 but these maternal hybrids rarely show any evidence of segregation 

 when used for crossing, and usually on self-fertilization breed true to 

 type. 



' N. Heribert-Nilssou. Spdltiiiti/dersfJifiiuiihien der (Jiiiulheni Laiiuuckiaim , 1915. 



