i9iS] 



BARTLETT— MUTATION IN OENOTHERA 



"3 



Turning again to table XII, it is seen that the progenies might 

 possibly be assembled in two groups, those with a mutation ratio 

 of approximately 400: i (group I of table XIII), and those with a 

 ratio of about 250: i (group II of table XIII). Testing separately 

 the ratios from these groups (see table XIIl), we find that the num- 

 ber of individuals in group II is too small to establish a significant 

 deviation from the ratio for group I. (The difference is o . 18 ± o . 13 

 per cent). It is therefore impossible to demonstrate either that 

 the mutation ratio is or that it is not the same for all the progenies. 



TABLE XIII 



Test op the tttness of the mutation ratios to the nearest Mendelian 



RATIO (235:1) 



If we assimie that it is justifiable to treat all of the progenies as one 

 group, the numbers are then large enough to indicate, not however 

 without considerable doubt, that the frequency of occurrence 

 of mut. nummularia is not in accord with the tetrahybrid ratio 

 255:1, but with some ratio l3ang between 330:1 and 450:1. Of 

 course we caimot assume that there is no mortality at all among 

 the seeds which produce mut. nummularia. If in the 30 per cent of 

 seeds 0. pratincola which never germinate even when fresh the mor- 

 tality among mutations and non-mutations were the same, then 

 the mutation ratio would not significantly deviate from 255:1. It 



