1908] CURRENT LITERATURE 67 
that not only variability itself, but the nature of the variation-curve and all of its 
“constants” are the product of the environment, and are a measure of the uni- 
formity or lack of uniformity of envi tal conditions. He concludes that there 
isno such thing as absolutely constant characters, and that the most constant may 
become quite variable under special conditions. There is no real distinction be- 
tween continuous and discontinuous variability, if the question of heredity is left out 
of account, as both of these as well as transitions between them may be induced 
by changes of the environment. The applicability of QUETELET’s law, which seems 
to indicate that variability is due to chance, depends upon the fact that the differ- 
ent values of the variable characters are determined by the coordinate action of 
several independently variable environmental factors. In an attempt to analyze 
the effective factors in the environment it is shown that the values of variable 
characters of Sedum increase directly as the quantity of carbohydrates increases 
and inversely as the quantities of available water and of salts increase. Chemical 
analysis of plants grown in the different habitats are presented in support of this 
W. 
PEARL" has studied the variability of Chilomonas paramecium and Parame- 
cum caudatum living under favorable and unfavorable conditions. He finds that 
the types of the two populations are significantly different, that of the less favored 
culture being smaller and relatively more slender. The variability was the same 
in the two cultures; but the curve of the well-fed culture was positively skew, 
while the poorly fed presented a nearly normal variation-curve. There is a 
correlation of 0.6 between length and thickness, and also a significant correlation 
tween size and form, which is recognized as opposing DrrescH’s statement 
that proportionality is absolutely independent of size. In Paramecium, PEARL?° 
also finds that there is a high degree of correlation or ‘“‘assortative mating” be- 
tween the two members of a pair of conjugating Paramecium, the coefficient 
Hs from 0.43 to °.79 in respect to length, and 0.217 to 0.349 in respect to 
readth. : 
the length. It is also found that the conjugants are differentiated as a c 
from the non-conjugating population, the former being smaller and less variable. 
From this the author concludes that conjugation tends to restrict rather than 
imctease variability, and that the conjugant type bears much the same relation to 
7 hon-conjugants as that borne by the germ-cells of a metazoan to its soma. 
” evolutionary progress must rest upon changes in the conjugant type. 
Of interest for the purely methodological side of biometry is a brief paper 
tio "9 Peart, R., Variation in Chilomonas under favorable and unfavorable condi- 
ms. $233-72. figs. 7. O. 1906. 
: *° Peart, R., A biometrical study of conjugation in Paramecium. Biometrika 
5213-297. figs. 9. F, 1907. 
