iQis] 



BARTLETT—MASS MUTATION 



429 



rigorous experimental methods, might have been overlooked the 

 year before. The results from the new Fi cultures of 1914 are set 

 forth in table I. 



TABLE I 



Analysis of the Fi seedling cultures op Lexington E 



* Indicates seeds from one capsule. 



t 25 plants of f. typica from culture 2, grown to maturity, were imiform. The remaining plants of 

 f . typica were discarded in the rosette stage. 



J Indicates that the mutation was grown to maturity. 



It is clear from table I that the Fi did not point to Lexington E 

 as a specially mutant strain. There were only 5 mutations in a 

 progeny of 522 plants. Moreover, 2 of the 4 types obtained, mut. 

 latifoUa and mut. graminea, were common to the other strains of 

 O. pratincola. 



In 1914, F2 progenies were grown from 3 plants of 0. pratincola 

 f. typica belonging to strain E, and the progeny of a fourth was 

 grown in 1915. The results are summarized in table II. 



The F2 shows a decidedly greater degree of variability than the 

 Fi. One progeny only, that from Lexington E-5, shows mutations 

 in excess of the number of typical plants; the other three progenies 

 indicate a degree of mutabiUty more comparable with that of 

 certain derivatives of 0. Lamarckiana, such as 0. scintillans. The 

 Fi, however, was, if anything, less mutable than 0. Lamarckiana 

 itself. Successive generations seem to show an increasing degree 

 of mutability. Only one F3 progeny from f . typica has been studied. 

 The parent belonged to the progeny of Lexington E-5, that is, it 

 was selected from the most mutable hne. The analysis of the F3 

 culture is shown in table III. 



The salient fact shown by the data for the Fi, F2, and F3 pro- 

 genies is that the number of mutations varies inversely with the 



