112 BOTANICAL GAZETTE [February 



were first found in 0. pratincola. Hunger^' has recently recorded 

 observations on selective mortality in the seeds of O. LamarcMana 

 which can only be interpreted as showing that the mutations of 

 this species have decidedly a greater survival value than the 

 parent form. 



It is often remarked that the Onagras are not most usually 

 foimd in undisturbed habitats with other native plants, but rather 

 as weeds in fields and waste places, among the aliens of our flora. 

 Wherever the soil is disturbed, as by plowing, road-making, 

 excavating, they are frequently foimd in large nimibers. They 

 often dominate the flora on made land and on new railroad embank- 

 ments, but are for the most part replaced by other .weeds when the 

 soil ceases to be disturbed at intervals. A fallow field which con- 

 tains many Onagras for a season or two after cultivation is dis- 

 continued will thereafter contain fewer each year. If again plowed, 

 it will apparently be restocked by the germination of seeds 

 which have lain dormant, perhaps for years. Selective mortality 

 among dormant seeds might result in such a field being restocked 

 with plants among which mutations would be unexpectedly 

 nimierous. 



The most interesting fact shown by table XII is that the fre- 

 quency of mut. nummularia cannot correspond with any Mendelian 

 ratio except that of a tetrahybrid splitting in the ratio 255:1. 

 In the case of a number of progenies, to be sure, the ratio of muta- 

 tions to plants more nearly approximates the trihybrid ratio 63 : i, 

 but it has already been shown that in each such instance the high 

 mutation ratio is associated with a low percentage of germination. 

 When the ratio of mutations to seeds is dealt with, there is no case 

 of an approximation to the 63:1 ratio. The data of table XII, 

 recalculated, are stated in table XIII in such form as to show that 

 no single progeny was large enough to prove that the 255:1 ratio 

 might not be the true one. On the contrary, the data afford no 

 reason to believe that the mutation ratio is 255:1. It may be 

 because of the smaUness of the cultures that no single progeny 

 shows a significant deviation from this ratio. 



" Hunger, F. W. T., Recherches experimentales sur la mutation chez Oenothera 

 Lamarchiana, execut&s sous les tropiques. Ann. Jard. Buitenzorg 27:92-113. 1913. 



