1919] Frost: Mutation in Matthiola 105 



and the other abnormal plants presumably gives a better comparison 

 as to mean and variability, but the conclusion is the same in either 

 case. The three many-noded (late) parents descended from "WG9- 

 ClO give no definite indication of being genetically different from the 

 "check" lots not descended from WGQ-CIO, while the variability con- 

 stants are sufficient, taken alone, to make probable the genetic differ- 

 entiation of the fewer-noded progeny of WG9-C10. Apparently all 

 the fewer-noded progeny of WG9-C10 that were tested — seven, when 

 WG9-C10-C10, a crenate-leaved apparent mutant (tables 12 and 13), 

 is included — w^ere either simplex or duplex for presence of an earliness 

 factor or factors. 



The variability of all the thirty progeny lots, taken together, is 

 high, as might be expected, though decidedly below that of the progeny 

 of early parents. This high variability is due only in very small part 

 to the progeny of the five or six supposedly mutant parents; the last 

 thirteen lots, alone, are much less variable than the mixed early lots. 

 The portion of the cultures containing these progeny lots from 

 aberrant parents was conspicuous for irregularity of germination, and, 

 on the whole, a relatively low^ rate of germination. 



A few of the last thirteen lots give more evidence bearing on the 

 origin of WG9-C10. The early WG9-syn3-M10 (tables 12 and 13) 

 gives no evidence of genotypic differentiation from its ordinary sib, 

 WG9-syn3-Mll ; WSI-W2I6, another phenotypically early parent, 

 also failed to transmit earliness to its progeny. CG2-C2-C6, on the 

 other hand, although itself an ordinary plant, shows a rather sus- 

 picious tendency to the production of early and few-noded progeny, 

 but better evidence would be required for any positive conclusion. 

 WG9-C10-C10 appears, from the data in tables 12 and 13 and from 

 observation of the flowering of plants of the next generation in the 

 1911H cultures, to have been heterozygous for the early type, as well 

 as for the crenate-leaved type. We find in this test no definite indi- 

 cation that the early type has appeared elsewhere than in WG9-C10 

 and its descendants. 



The F2 progeny of WG9-C1, an abnormal plant whose F^ progeny 

 were unusually and uniformly early but not few-noded, have been 

 included with the other check lots without question. This treatment 

 seems justified by the flowering data, which do not indicate any 

 repetition of the precocious development of the first-generation plants ; 

 the peculiarities of the F^ cultures, if not a mere cultural accident, 

 presumably depended on the very abnormal development of the parent, 



