444 



BOTANICAL GAZETTE 



[DECEMBER 



obtained from 15 capsules of the primary mutation, Lexington E-5- 

 238. The progeny consisted of 25 plants, only 16 of which survived 

 transplanting from the seed pan. None of the progeny resembled 

 the parent. All were extreme dwarfs which resembled, but were 

 not identical with, mut. rewluta and mut. setacea. They differed 

 mainly in the thicker leaves, which in 4 plants were narrow but 

 not markedly revolute. Although a very nondescript lot, differing 

 much among themselves, 5 most resembled mut. revoluta, and 7 

 mut. setacea. The result of this culture might almost have been 

 predicted. The mass mutability was inherited by mut. gigas from 

 f. typica. In view of the dependence of the gigas characters upon 



TABLE V 

 Analysis of an Fi cttltuke prom self-pollinated mut. latifolia, Lexington E-36 



The parent plant belonged to the Fi from Lexington E, f. typica; see table I, culture 2, 

 for position in pedigree 



* Seeds from 7 capsules, containing respectively 54, 72, ^g, 30, 38, 70, and 62 seeds. 



t Ncs. 32 and 43, a new mutation; no. 51 mut. signs ( ?), morphologically identical with Lexington 

 E-s-238 which had 28 chromosomes; see table II, culture s from Lexington E-s, for position of latter 

 plant in pedigree. 



X Excluding, of course, mut. latifolia. 



the double complement of chromosomes, which would in general 

 be handed on to any secondary mutations, it follows that the muta- 

 tions occurring en masse would not be identical with those from 

 f. typica. Furthermore, chance irregularities in chromosome 

 distribution might increase the polymorphism of the progeny. 

 In such a highly modified germ plasm irregularities would be 

 expected. 



The cross mut. gigasXi. typica yielded i6o seeds in a single 

 capsule, of which lo germinated. The plants were all extreme 

 dwarfs, of the most nondescript nature, hardly any two alike. All 

 had thick leaves, some plane, others revolute. No mutation of 



