164 



GENETICS 



[BoT. Absts., Vol. X, 



be a more reliable measure of variation than the probable error where only a small number of 

 plots is used. Data from 2 extensive experiments are given. The necessity of making pre- 

 liminary experiments before planning extensive trials is pointed out, and the need of care in 

 the choice of experimental plots is emphasized. [See also Bot. Absts. 10, Entry 1037.] — Carl D. 

 La Rue. 



1069. March, Lucien. La methode statistique. [The statistical method.] Metron 1: 

 22-52. 1920. — The aim of science is the classification of observations. The experimental 

 method studies the relation of an effect to a single cause when other causes are held constant. 

 When the various causes can not be isolated the statistical method must be used. Analysis 

 of this method may be divided into 3 parts: (1) Comparison of centering constants; (2) study 

 of variability within the group; (3) study of relationships between groups. Properties of the 

 median and arithmetical mean are discussed. — John Rice Miner. 



1070. Marshall, Roy E. Report of three years' results in plum pollination in Oregon. 

 Proc. Amer. Soc. Hort. Sci. 16: 42^9. 2 pi. 1919 [1920].— See Bot. Absts. 6, Entry 129. 



1071. Mellon, Ralph R. The life-cycle changes of the so-called Corynebacterium 

 Hodgkini, and their relation to the mutation changes in this species. Fourth paper on diph- 

 therioids. Jour. Med. Res. 42: 61-76. 1920. — The strain of C. Hodgkini used, when culti- 

 vated under ordinary conditions, is generally non-granular. It infrequently develops "giant 

 cocci," which respond to the stimulus of new environment; e.g., when cultured in broth plus 

 rabbit serum they give rise to coccoid forms of irregular size, which continue development, 

 with increase of chromatin, when 1 per cent maltose is added to this medium. Whether this 

 process is degenerative or vitalistic is seen in the stabilization of the offspring obtained by the 

 different modes of reproduction, viz., gemmation and sporulation (?); gemmation produced 3 

 generations unchanged. Response to environment is further emphasized by the fact that 

 granular coccoid forms develop in serum broth but do not develop in hydrocele fluid. Ar- 

 rested bacillary growth favors coccoid formation, a phenomenon which is constant. — The 

 staining of chromatin and achromatic areas of the organism follows laws analogous to chemical 

 reversibility and tautomerism. — The morphologic mutation effected by rotation seems to be 

 of a cyclic character; however a return of the coccoid to the bacillary form has not been 

 determined. — Andrew I. Dawson. 



1072. MoL, W. E. DE. Over het voorkomen van heteroploide varieteiten van Hyacinthus 

 orientalis L. in de HoUandsche kulturen. [On the occurrence of heteroploid varieties of Hya- 

 cinthus orientalis in Dutch cultures.] Genetics 3 : 97-192. 1921. — The named hyacinth clones 

 in which the somatic numbers of chromosomes were counted may be classed as follows: 



-John Belling. 



