522 THE AMERICAN NATURALIST.  [VoL. XXXIII. 
curves agreeing with any observed distribution of varying individuals. 
To solve some of the equations, logarithms, trigonometric functions 
and gamma functions have to be employed, and it is too much to 
hope that a large proportion of naturalists can use even these simple 
methods. Yet naturalists who desire to reach best results from their 
data must get out their algebras, trigonometries, and logarithmic 
tables, and brush up on these subjects. Formulas are used very 
freely, which is no doubt an offense to some. It should be remem- 
bered, however, that a formula is only an abbreviated statement of 
operations, and very useful to any one who knows how the operations 
indicated are performed. The author gives the various indices of 
variability. He objects to the use of the coefficient of variability 
(obtained by dividing the index of variability by the mean) and 
maintains that the view that we expect high indices of variability 
where the mean is high is untenable. As evidence for this con- 
clusion he points to the fact that in the four numerical examples 
given in his paper the index of variability is inversely propor- 
tional to the mean. This is, however, absolutely no evidence 
against the value of the coefficient of variation, since, doubtless, 
the large characteristic which has a small index of variability is, 
in comparison with the small character having a large index, 
relatively even less variable than a comparison of the mere indices 
would indicate. 
Correlation is treated of briefly, and an excellent short method of 
calculating the coefficient of correlation is given. The calculation of 
spurious correlation of indices is not given — an important omission ; 
and little is said about multimodal curves, which are si tat in 
biological statistics. 
A bibliography of 111 titles is given (from which Fechner’s 1897 
work is omitted), and tables of formulas add greatly to the useful- 
ness of the paper. 
Altogether the paper is one which every student of statistical biol- 
ogy —and this ought to include every systematic worker in zodlogy and 
botany — should have. 
Inheritance of Acquired Qualities.'\—Conidia of Aspergillus 
niger of common origin are reared (4) in Raulin solution; (4) 
in Raulin solution + 6 per cent NaCl for ove generation ; and (C) in 
Raulin solution + 6 per cent NaCl for zwo generations. Then: 
1 Errera, L. Heérédité d’un caractère acquis chez un champignon pluricellulaire, 
Bull. Acad. Roy. de Belg., 1899, pp. 81-102. 
