136 EXPERIMENT STATION RECORD. 



not known as a native species in any part of the world and that there is reason 

 for believing that the plant has come down to us as a hybrid and a product of 

 a lone: period of cultivation. 



The variability of Bacillus solanacearum, J. A. Honing (Meded. Deli 

 Proefstat. Medan, 7 (1912), No. 5, pp. 196-208; Centhl. Bakt. [etc.], 2. Abt., 36 

 (1913), No. 19-25, pp. ^91-499) .—The author, finding B. solanacearum very 

 variable in its characters, experimented with a great number of cultures in 

 various nutritive media. 



As a result, it is stated that with a large number of different carbon and 

 nitrogen sources this fungus showed development only in a part of the cultures 

 employed. Differences of alkalinity exert little, and those of light and tem- 

 perature no, influence on the variability. Only a few individuals among many 

 thousands are able to adapt themselves to new substrata, and the acquired 

 modifications are not sufficiently constant to rank as mutations. It is stated 

 also that in a few instances Bacterium deliense n. sp. has shown the same 

 tendency to vaiy. 



Constant variants of Capsella, H. Hus (Ahs. in Science, n. ser., 31 (1913), 

 No. 949, p. 375). — In the investigations of the author pedigree cultures from 

 Capsella showed the existence of a number of constant forms that have not 

 previously been reported, some of which are not identical with the biotypes 

 described by Shull. Emphasis is laid by the author on the importance of the 

 study of seedling stages in the identification of these forms. 



Preliminary note on the inheritance of self-sterility in Reseda odorata, 

 R. H. CoMPTON (Proc. Camjyridge Phil. Soc, 17 (1912), No. 1, p. 7).— A brief 

 summary is given of the author's investigations relating to the self-sterility 

 of mignonette, some forms of which are believed to be self-sterile while others 

 are completely self-fertile. Seed was obtained from various sources and plants 

 grown, and the observations on the first generation are given, which seem to 

 confirm the hypothesis regarding sterility. 



Self-sterile plants when bre<l among themselves gave self-sterile offspring 

 only, indicating that self-sterility is recessive, according to the Mendelian 

 theory. Some of the self-fertile plants when self-fertilized yielded self-fertile 

 offspring, and the same result was obtained when crossed with self-sterile 

 plants. These are regarded as homozygous. Other self-fertile plants when 

 self-fertilized yielded approximately three self-fertile to one self-sterile offspring. 

 When crossed with self-fertile plants about half the progeny were self-fertile, 

 the other half being self-sterile. 



Other characters were noted in the study, and it appears that tall and 

 dwarf races exist in mignonette and that the Fi between them is intermediate 

 in height. Differences in pollen were observed, the orange red color appearing 

 to be a simple dominant to bright yellow and self-fertilized heterozygotes 

 yielding about three reds to one yellow. 



Some relations between root characters, g'round water, and species distri- 

 bution, W. A. Cannon (Scicvce, n. ser., 37 (1913), No. 950, pp. Ii20-J,23) .—The 

 author reports observations on the root habits of desert shrubs and trees, dis- 

 cussing the relations between the root habit, the ground water, and species dis- 

 tribution. 



He claims that there is an intimate relation between the occurrence of certain 

 species of trees and the character of their roots with regard to the depth at 

 which perennial water may be found. Among the illustrations cited he states 

 that the mesquite, which is the most widely distributed tree of the Tucson 

 region, occurring not only on the flood plains of streams but also at higher 

 elevations, assumes a tree form only where the roots can reach to the water 

 table. Elsewhere the character of the plant is that of a shrub. From a com- 



