820 EXPEBIMENT STATION EECOBD. [Vol. SS 



logical and cultural differences which are permanent, yet in other characteristics 

 they are so alike that it is considered best to regard the adapted forms as 

 varieties of a single species. The nodules occurring on Ceanothus, Cycas, Alnus, 

 and Myrica are said not to be caused by P. radicicola, those on Ceanothus differ- 

 ing morphologically from those found on the LeguminosjE. The authors do not 

 consider conclusive the evidence that the nodules on Elaeagnus and Podocarpus 

 are caused by legume bacteria, nor is the proof conclusive that any of these 

 nonlegume plants are concerned In the fixation of atmospheric nitnjgeu. 



Extensive experiments were conducted in an attempt to infect nonlegume 

 plants with nodule bacteria, always with negative results, and it is claimed 

 that no conclusion can be drawn as to the possibility or probability of develop- 

 ing or finding nodule bacteria that will grow on nonlegume plants. 



An extensive bibliography of the subject is appended. 



The abstract is by A. L. Whiting. 



The behavior of self -sterile plants, E. M. East (Abs. in Science, n. ser., 46 

 (1917), No. 118S, pp. 221, 222).— According to the author, there are two problems 

 connected with the inheritance of self-sterility in plants. One is the relation 

 between self-sterile and self-fertile plants, the other the behavior of self- 

 sterile plants when crossed together. In Nicotiana self- fertility is com- 

 pletely dominant over self-sterility. Either of the self-sterile species ^V. alata 

 or N. forgctiana may be crossed with the self-fertile species A', langndorffii, 

 resulting in each case In an Fi generation which Is completely self-fertile. 

 The F> plants are said to show the usual monohybrid ratio of 3 self-fertile to 1 

 self-sterile. 



Discussing the results obtained in his investigations, the author concludes 

 that«the behavior of self-sterile plants in intercrosses is regulated by several 

 transmissible factors, all of which are distinct from the single basic factor for 

 self-sterility and which presumably may be carried by self-sterile plants. A 

 plant homozygous for self-sterility can neither be fertilized by Its own gametes 

 nor by the gametes of any other self-sterile plant of like constitution as regards 

 these regulation factors, but any two plants differing in these regulatory 

 factors are cross-fertile. 



Twin hybrids from (Enothera lamarckiana and CE. franciscana when 

 crossed with CE. pycnocarpa, (J. F. .\tki.nson (.Ms. in Science, n. ser., ^6 (1911), 

 No. 118S, p. 222). — According to the author, when CE. lamarckiana is crosseil 

 with (E. pycnocarpa, there is a splitting In the F, generation with the production 

 of twin hybrids. One of the twins is characterized by the pycnocarpa type, 

 while the other Is designated as a latiMrckiana tyi>e. These twin types are 

 fixed In the first generation, and are repeated In the Fj and probably in the 

 following generations in accordance with the usual behavior of twin hybrids 

 determined by De Vries. 



When (E. franciscana and CE. pycnocarpa are crossed, there is said to be a 

 splitting in the Fi with the production of twin hybrids, and in the Fj genera- 

 tion there is a one-sided splitting similar to that which occurs in the Fi of 

 twins from CE. hookcri and Q'J. lamarckiana, as described by De Vries. In this 

 second grouping the pyrnocarpa type twin is said to have a hybrid constitution, 

 while the franci.tcana type of this generation is fixed in the Fi and repeats itself 

 Id the Fi. The franciscana twin Is believed to carry the pycnocarpa factors, 

 but In a subordinate or permanently latent condition. 



Najning American hybrid oaks, W. Trkt.eask (.\bs. in Science, n. ser., ^6 

 (1917), No. I1S4, p. 244). — I" a study of American oaks, the author has Investi- 

 gated 38 known or probable hybrids among tlie oaks of the Ignited States. To 

 the number of accepted hybrids already recorded two are added in this pai>er. 



