432 EXPERIMENT STATION EECOBD. [VoU87 



a much longer exposure to fresh water is considered significant. Suggestions 

 are made that some protoplasm may part more readily than other with certain 

 salts with which it may be chemically or mechanically united, and that the less 

 tolerant protoplasm may consist more largely of such substances as globulins 

 or other colloids, which undergo a change of state as soon as the concentration 

 of the salts falls below a certain limit. 



Copper in the flora of a copper-tailing region, W. G. Bateman and L. S. 

 Wells (Jour. Amer. Chem. Soc, S9 {1917), No. 4, pp. 811-819).— Plants grow- 

 ing in a copper-tailing region were found to contain copper, arsenic, antimony 

 and zinc, the amounts of copper ranging from 0.0046 to 0.G21 per cent. Larger 

 quantities were present in the dead than in the living tissue and in the bark 

 than in the other parts of the plant. 



It is noted that " many plants are unable to adapt themselves to this novel 

 factor of environment, while others flourish in spite of it, thus showing a 

 decided selective activity." 



Models to illustrate segregation and combination of Mendelian characters, 

 H. M. Chibber {Agr. Jour. India, Itidian Set. Cong. No., 1916, pp. S0-S5, figs. 

 2). — This paper describes a scheme for demonstrating the behavior of hereditary 

 characters that conform with the Mendelian principles. The dominant deter- 

 miners are represented by colored beads, while the recessive determiners are 

 repre.sented by beads identical in size and shape but colorless. The method of 

 demon.strating the composition of diflerent zygotes and of their gametes is illus- 

 trated with reference to one, two. and three characters, respectively. The 

 demonstration of the phenotyplc composition Is also illustrated. 



The probable error of a Mendelian class frequency, R. Peakl (-4>ner. Nat., 

 51 (1917), No. 60S, pp. 144-156). — In view of the growing need for adequate 

 and clearly understood tests for the statistical significance of differences be- 

 tween expectation and observetl results in Mendelian experimentation, the 

 author offers a method supplenientary to those in use of calculating and ex- 

 pressing the error.s due to random sampling, of a Mendelian class frequency. 

 The method consists essentially in expressing each expected frequency as the 

 prubable quartile limits, determined from the ordinates of a hypergeometrical 

 series, for that class frequency in a supposed second sample of the same size as 

 the observed sample drawn from the same population. Simplifications of 

 method are suggested and illustrated. 



Note on the inheritance of crossability, W. O. Backhouse {Jour. Genetics. 

 6 {1916), No. 2, pp. 91-94). — The author found that in the progeny of a hybrid 

 wheat crossed with rye there was a segregation of crossability. The data here 

 obtained from this and other crossings described are considered to justify 

 caution in ascribing importance to the production and fertility of hybrids as 

 a guide to relationship. It Is considered as very probable that the deciding 

 factor as to whether two individuals will hybridize or not may depend upon 

 one simple Mendelian factor. 



The results of hybridization in Salix, S. Ikeno {Bot. Hag. [Toki/o], SO 

 {1916), No. 557, pp. 5/6-520).— Experimentation, including artificial hybridiza- 

 tion since 1910 between S. purpurea miiltinervis and S. purpurea gracilistyln, 

 is said to have shown that the hybrids are not always constant. A certain 

 amount of splitting was revealed, somewhat obscured by the failure of some 

 of the plants to live, so that the question as to Mendelian behavior in this 

 respect is not yet settled. 



Evolution by means of hybridization, J. P. Lotsy {The Hague: Mariinus 

 Nijhoff, 1916, pp. X+166, figs. 2). — Discussing some evidences and views which 

 have been offered regarding the problems and methods of organic evolution. 

 the author claims that crossing is the cause of the origin of new types which 



