iQiS] BARTLETT— MUTATION IN OENOTHERA 115 



are infinitesimal that all 9 would be included among 11 plants 

 chosen at random. 



3. A tetrahybrid might give as many as 16 phaeno types in the 

 F2. All of these would have a greater frequency than 1:255 

 except the pure recessive. We have seen that mut. nummularia 

 cannot have a greater frequency than 1:255, ^.nd have also seen 

 that it is not a pure recessive, for in the next generation after it 

 originates it gives rise to several distinct types. 



4. In the case of one F2 progeny (Lexington C-91, see table III) 

 from a single mother plant, 1,539 seeds from 6 capsules gave 3 

 specimens of mut. nummularia, whereas 337 seeds from one capsule 

 gave 5. Such a result shows a frequency varying from 1:60 

 to 1:5x3 on capsules from the same spike. From a Mendelian 

 standpoint it is practically impossible to explain such a result. 



The mutation phenomenon in 0. pratincola cannot be explained 

 away by any reasonably plausible stretching of Mendelian theory. 

 On the contrary, it seems obvious that mutation is quite a different 

 process from hybrid segregation, although both processes may 

 occur simultaneously. 



Mut. nummularia is the only one of the mutations of 0. pratin- 

 cola the frequency of which has been determined. None of the 

 others Jias been observed throughout the complete cycle from seed 

 to seed and carried into a second generation. In tables II-XI all of 

 the variants except mut. nummularia are thrown together as 

 "other mutations or suspected mutations." In explanation of this 

 mixed category, it is necessary to state that all unusually small or 

 unusually large plants, regardless of whether or not they appeared 

 otherwise different from the mass of the culture, were coimted as 

 "suspected mutations," in order to be sure that no such variation 

 as a mut. nanella would be passed over. To judge from past 

 experience, most of the "suspected mutations" will develop as 

 quite normal plants. Consequently the mutability of O. pratincola 

 is probably not as great as might be assumed from the tables. 



Mut. Mummularia a discontinuous variation 



Critics of De Vrees' work on mutation in Oenothera Lamarckiana 

 have not infrequently expressed skepticism as to whether or not the 



