166 Genetic Behmrionr of the Hyhrkl Primula Kewensis 



of fluribunda var. isahellina was crossed with six Kewensis fathers, and 

 in four of these crosses maternal hj'brids were obtained ; the two other 

 crosses failed. Two of the maternal hybrids thus obtained were crossed 

 again with Keivensis, and one of them gave maternal hybrids. On the 

 other hand there is a case in which a Kewensis plant was fertilized by 

 four verticillata fathers, one cross only gave offspring, the rest failed. 



From 91 crosses of verticillata x floribunda and reciprocal, Kewensis 

 appeared only twice. In each case one plant only was obtained and 

 both these plants were fully fertile and bred true to Kewensis. The 

 chromosomes of one hybrid were counted by Miss Digby, who found 

 the haploid number to be 18 and the diploid number 36. We have 

 never obtained the diploid original form of Kewensis as a result of our 

 crosses. We may note here that floribunda and verticillata have never 

 reappeared from the self-fertilization of Keivensis. 



Of 28 plants of verticillata used as females in crosses with floribunda 

 and Keivensis, 7 gave maternal hybrids. Of 17 plants of verticillata 

 tested for parthenogenesis, one only gave viable seed. Of floribunda, 

 39 plants were used as females in crossing, 9 gave maternal hybrids. 

 Ten plants of floribunda were tested for parthenogenesis, and gave no 

 seed. Of Kewensis, 34 plants were used as females for crossing, and 

 11 gave maternal hybrids. Twelve plants of Kewensis were tested for 

 parthenogenesis ; 3 gave viable seed. 



Tested by the viability of the seeds there is no great difference 

 between the results of crossing and those obtained in the trials for 

 parthenogenesis; but judged by their appearance there is a good deal 

 of difference. In the crosses, the seed is often described as fair or 

 good, whereas m the parthenogenesis tests onl}^ once from a verticillata 

 did we obtain any seed described as good, and once seed described as 

 fair fi'om Kewensis. 



Taking into consideration that the plants were in a glass cage, 

 and the irregularities previously mentioned, the possibility that the 

 "maternal hybrids" are parthenogenetic in origin cannot be said to 

 be removed. The question must for the present be left open. We are 

 continuing the experiments this year'. 



1 Perhaps in favour of the view that these offspring are parthenogenetic is the fact that 

 using the pollen of P. Forrestii, a species far removed from this group, maternal offspring 

 were obtained from rerticiUata. 



