466 BOTANICAL GAZETTE [DECEMBER 
his students. Much of the book has been rewritten and many additions of new 
examples and new methods are given, making it more indispensable than ever 
as a handbook for the student of this important phase of biology. 
The changes which have been made are too numerous to be considered in 
detail, but the most noticeable are as follows. (a) The section is omitted which 
dealt with the quantitatve expression of terms used by botanical taxonomists 
in the description of leaf-form. (6) The subject of correlation has received new 
and altogether better treatment by the substitution of YuLrs’s method for 
DUNCKER’S, and the addition of PEARSON’s method for determining the correla- 
tion between qualities not quantitatively measurable. (c) Two additional 
types of asymmetrical curves are presented. (d) A section is introduced dealing 
with MENDEL’s law of alternative inheritance. (¢) A 22-page chapter on the 
results of statistical biological study is substituted for the 2-page chapter on the 
applications of statistical methods. (f) A professedly complete bibliography is 
given instead of a selected one. 
A comparison of the bibliography with that given in the first edition shows 
in an interesting way the remarkable activity which has developed in this field. 
In the earlier edition thirty-nine titles were given, in the present edition there 
are 265 references, 186 of which bear dates later than the date of publication of 
the first edition. As is usual in extensive bibliographies the attempt at com- 
pleteness leaves something to be desired. A number of titles not found in the 
list occur to the reviewer as being of more value statistically than some which 
are given. 
The unique feature of the chapter on the results of statistical studies is a 
tabulated analysis of the literature, showing the general bearing, and in some 
instances the point of view, of each paper. At least two of these papers are listed 
under subjects to which they make no significant contribution; ¢. g., HarsH- 
BERGER on “The limits of variation in plants” and KELLERMAN on “Variation 
in Syndesmon thalictroides” are classed as dealing with correlation, but neither 
paper treats specifically of questions of correlation, and the data given by each 
are too meager to be of value to students who would be interested in turning them 
to account in the study of correlations. 
Everyone will appreciate-how difficult it must be to keep free from errors 4 
work made up so largely of tabulations. A reference on p. 113 to Table X means 
Table X of the first edition, which has become Table XII of the present edition. 
Botanists will be astonished to see Syndesmon thalictroides classed as a desmid 
on p. 78. 
On the whole, the second edition is a very marked advance over the first, 
and there is every reason to expect that with its assistance the bibliography of 
statistical biology will rapidly advance in the coming years, not alone in bye 
number of titles but also in the clearness and completeness of mathematical 
analysis and in the importance of the conclusions reached.—GroRGE H. SHULL. 
