2 Contributions from the Gray Herbarium 
South America. Many new species were described, the range of 
many others extended, and for the first time a certain amount of 
order was brought into the chaos which had hitherto reigned in the 
classification of the Middle American species. In the same year 
Hemsley recorded in the Biologia Centrali-Americana 39 species, 
besides one unidentified species, from Mexico and Central America. 
This list is based chiefly on Bennett’s work, but many of the identi- 
fications there recorded are erroneous. Between this date and 1891 
several new species of considerable interest were published by Wat- 
son, chiefly from Palmer’s and Pringle’s collections. 
A new epoch in the study of the genus Polygala was initiated by 
the appearance of Chodat’s monograph in 1893. Here the species 
of the world, some 405 in number (of which some 57 were recorded 
from Mexico, Central America, and the West Indies), were divided 
into ten well-marked sections, and the great value of the characters 
to be derived from the persistence or deciduousness of sepals and 
wings, the nature of aril, capsule, and style, and the clothing of the 
seed, received due recognition and illustration. Unfortunately this 
work, on which all future taxonomic study of the genus must be 
based, is marred by many typographical and other inaccuracies. 
One finds two new species described under the same name, and even 
two consecutive valid species maintained with the same trivial; un- 
fortunate errors of geography, of author-citation, and of nomen- 
clature; omission of mention of species described in easily accessible 
works; and keys so loosely drawn up as to be frequently mislead- 
ing. All these faults, however, cannot obscure the real worth of 
Chodat’s monograph, which in its main lines is thoroughly sound, 
and from its wealth of illustration must always remain an indis- 
pensible work of reference. 
The genus Polygala, in the wide sense in which it is taken by 
Chodat and by the writer, includes forms of great diversity in habit 
and considerable difference in structural characters, but the ten sec- 
tions into which it is divided by Chodat, although well defined, are 
so regularly graduated that the separation of any one of the few 
(Phlebotaenia, Badiera, Acanthocladus, Chamaebuxus) sometimes 
recognized as distinct would involve in consistency the recognition 
of all. Its two or (if the species of Chamaebuxus with beaked keel 
be distinguished) three primary divisions based on the structure of 
the keel possess greatest claim to generic rank, although no one has 
