114 GENETICS [Bot. Absts., Vol. IV, 



of psychical characteristics. — A group of 91 pairs of siblings from the orphan asylums of 

 California has been tested by means of the Stanford Revision of the Binet-Simon method 

 by Dr. Kate Gordon. The reason for selecting siblings from these orphan homes is because 

 the environmental training has been the same for both, thus eliminating that factor of dispute 

 of differential environment. By the table prepared from these results, a mean intelligence 

 quotient (I. Q.) of 92.857 (or 93) =*= 0.836 and a correlation of intelligence between the siblings 

 of 0.50S =•= 0.0524 are found. — The above results are compared with a table of data computed 

 from the examination of 2801 pairs of siblings from all parts of England and all stations of life, 

 by the "broad-category" method. The unit of this method is called the "mentace." Com- 

 parison of the two tables on equal basis was made by evaluating the mentace in terms of the 

 I. Q. unit. A mentace is 1/100 part of the range which limits the category "Intelligent," or 

 0.1604 I. Q. units. — The results of the second table show a correlation of 0.515 which agrees 

 excellently with the correlation of 0.508 of the first table. — Average genius is 50 I. Q. units 

 above the mean I. Q. of 93 and the average mental defective is 50 I. Q. units below the mean 



1. Q. Thus also a person with absolute-zero intelligence would be exceedingly difficult to 

 find, for even a low form of idiocy has a slight measurable intelligence. It is quite evident from 

 the comparison of the above data that there is a hereditary factor in the intelligence-resem- 

 blance of siblings, and should dispel the prejudiced belief that this resemblance is due to 

 differential environment. — S. W. Prince. 



700. Pearson, Karl. On generalized Tchebycheff theorems in the mathematical theory 

 of statistics. Biometrika 12: 284-296. November, 1919. — Generalizes Tchebycheff 's theorem 



for single variate in form 



where P is chance of deviation being smaller than Xcr 



For two variates, probability of an observation falling within ellipse X 



*0 



where I a involves high product moments. The theorems are tested empirically and found of 

 little practical value, due both to labor involved and roughness of limits obtained. — Sylvia 

 Parker. 



701. Pexxyp acker, John Young. Observations in the beach plum, a study in plant varia- 

 tion. Contrib. Univ. Pennsylvania Bot. Lab. 4:231-269. PL 66-70. 1919.— The beach 

 plum, which occurs both along the seashore and inland, blossoms on the average about May 



2, although certain shore forms do so as late as May 15. The writer indicates the marked 

 variation in size, in mode of branching, in vigor of shoots, in time of appearance, in size and 

 in hairiness of leaves, as well as in their blossoming. The flowers are androdioecious, and this 

 evidently accounts for fruitless plants, others sparsely fruitful and still others full-bearing. 

 The fruit varies greatly in size, color and palatability. Pure yellow fruits were collected, as 

 well as red, purple, and large blue-black drupes. While the purple and blue-black fruits are 

 often rich in tannin, the yellow fruits are comparatively poor in this. The future develop- 

 ment of the fruit is considered promising. — John W. Harshbergcr. 



702. Pezard, A. Le conditionnement physiologique des caracteres sexuels secondaires 

 chez les oiseaux. Du role des glandes genitales. [The physiological conditioning of the sec- 

 ondary sexual characters in birds. The role of the genital glands.] Bull. Biol. France et 

 Belgique 52: 176 p. 1 pi., 79 fi<j. 1918.— A careful study of the problem from several angles, 

 which may be reviewed under two heads. (1) Confirmation of previous observations in 

 that the castrated cock retains the male characters except for failure of comb and wattles 

 to develop, and the loss of crowing and sexual instinct. Similar observations were extended 



