BIOLOGY, EXPERIMENTAL — DAVENPORT. 99 



the work planned with this variety was abandoned. Occasion was 

 taken, however, to point out in a paper presented before Section G, 

 American Association for the Advancement of Science, December 30, 

 1904, and afterward published in Torreya. that Johannsen's results 

 are in exact agreement with Gal ton's law, instead of being contrary 

 to it. The importance of Johannsen's results, however interpreted, 

 make it a matter of some regret that they were not elaborated by 

 more approved statistical methods. As a corollary to his conclusions, 

 every change in bean-weights capable of being segregated by selec- 

 tion within the individual pure line is of the nature of a mutation. To 

 satisfy the desire for a more appropriate elaboration and to investi- 

 gate the occurrence and behavior of these minor mutations, the study 

 is being repeated this year with Henderson's bush limas. This 

 variety has proved highly satisfactory, as it forms a bush of moderate 

 size and blooms and fruits well when inclcsed in bags of mosquito 

 netting, this means being adopted to insure the purity of the hered- 

 itary lines. 



The transmissibility of certain abnormalities has been tested 

 with the following results : Of eleven specimens of Rudbeckia hirta, 

 the entire progeny of a strongly fasciated individual, only one showed 

 evidence of fasciation, but the offspring of a fasciated Ambrosia 

 artemisicsfolia had slight fasciations in 8 of the 15 specimens. Com- 

 pound spikes of Plantago lanceolata were transmitted, fully developed, 

 to one of 26 offspring. Six of the other specimens showed abnor- 

 malities which may be related in character to the compound spike. 

 These consisted in the occasional production of scapes bearing one 

 or several leaves, or even a well-formed rosette, while axillary to 

 these leaves arose scapes terminated by single unbranched spikes. 

 Pla7itago major did not transmit in a single well-marked case, in a 

 progeny of more than 100 specimens, a rosette of leaves about the 

 base of the flower-bearing portion of the scapes. The possibility 

 is not excluded, however, in this case, nor in other cases in which a 

 parental character is wanting in the offspring, that the young are 

 heterozygotes in which the character in question is recessive. This 

 point can only be determined by the study of a second generation. 



In the common garden form of Hclia?it/ms anmcus, known as the 

 Russian sunflower, two characters have been chosen for the study 

 of inheritance. One of these characters, the greater or less degree 

 of bifission of the leaves at the second node above the cotyledons, is 

 obviously a progressive character, while the other character — the 

 branching habit— is, at least in a certain sense, atavistic or degress- 

 ive. Both of these characters prove to be rather strongly inherited, 



