6 Contributions from the Gray Herbarium 
slightly less than a third of the known species. They are herbs, 
annual or perennial, in one section suffrutescent at base, with per- 
sistent wings and sepals, the keel bearing an infra-apical variously 
lobed crest. In two African species (P. Petitiana A. Rich. and 
P. nilotica Chod.) of this subgenus there is, according to Chodat, 
no crest whatever. In P. paucifolia Willd., of our eastern states, 
a member of the subgenus Chamaebuxus, there is a strongly lobed 
fimbriate crista, and in various Old World species of Chamaebuxus 
there is a crest of varying complexity. 
The Middle American species of Orthopolygala divide naturally 
into two sections. In one, confined to Central America, Mexico, 
and the adjacent United States in North America (but occurring 
also in South America), the upper cell of the capsule is dehiscent 
between the narrow wings, while the slightly shorter lower cell is 
usually indehiscent and its seed generally unappendaged. In the 
other section, including the bulk of the species, the capsule is wing- 
less, equally 2-celled, and both cells are dehiscent. The species of 
the latter section can be divided into very satisfactory groups on 
characters of seed-pubescence and aril. In this as in the other sub- 
genera emphasis has been laid in the present revision on seed- 
rather than on stigma-characters as offering more easily observ- 
able and definable and quite as trustworthy characters for the dis- 
crimination of species. 
The present revision was begun and partially completed during 
the summer of 1915 at the British Museum and at Kew Herbarium, 
and my grateful thanks are due to Dr. A. B. Rendle and Dr. Otto 
Stapf for the opportunity to study the collections under their charge, 
including many types of the utmost importance. Through the kind- 
ness of Dr. Stapf fragments of several types were transmitted to the 
Gray Herbarium, where the revision has been brought to completion, 
and where the examination of an abundance of Mexican material has 
led to corrections or fundamental alterations in my conception of 
several species. While at the British Museum I was, through the 
kindness of the Honorable William Fawcett, enabled also to study 
the collection of Polygala i in the — - om Jamaica Botanic 
Garden. Grateful acl re also due Dr. B. L. Robinson 
for his interest and assistance in the cabieatiot of this revision. 
In nomenclature the International Code has been strictly fol- 
lowed. The herbaria in which the specimens cited are located have 
