Pammel — Anatomical Characters of Seeds of Leguminosae. 133 



called attention to the mucilaginous spiricles, somewhat simi- 

 lar to those in JRuellia, in some sections of the genus Euphor- 

 hia, and the total absence of mucilage in some species. In 

 some recent studies of the testa of Berberidaceae, I have 

 found some excellent generic characters; but very closely 

 related species it is often difficult to differentiate. Sub- 

 genera in Berheris are easily separated. In Cruci ferae (417) 

 I have also found some excellent characters. In studies of 

 this kind one is likely to fall into the same errors that the older 

 systematic botanists did, from not having had an abundance 

 of material to work over. Some botanists have drawn infer- 

 ences from a study of scant material and few species. 



In examining the seeds of Leguminosae , my studies have 

 been confined to genera and subgenera. Some excellent ana- 

 tomical characters have been found in tribes and some genera, 

 but it is often difficult to differentiate closely related species. 

 This statement, however, applies equally well to morphological 

 characters. Size is by no means always a good criterion in 

 anatomy, as Schumann (443) has shown, or in taxonomy. The 

 seed is, however, more constant under natural conditions than 

 are the vegetative organs. The general characters of seeds 

 often afford most valuable characters. These have been used 

 in many cases, as in Engelmann's work on Euphorbia and nu- 

 merous other systematic works. Quite recently Cunningham 

 (321) has proposed to use the seed characters of Plantago^ 

 and Wiegand the same for Galium, 



SYSTEMATIC. 



At first I contemplated only a study of the genera repre- 

 sented in Gray's Manual ( 6th ed.). As the work progressed, 

 however, it seemed best to include all of our economic genera 

 and a few additional from the South and Southwest. 



For synonymy, the Kew Index has been followed, as this 

 agrees essentially with the names applied to the species in 

 Gray's Manual (6th ed.). In the ganxx^ Phaseolus I have 

 followed Watson. Where these two systems disagree the 

 synonymy is placed in parenthesis. Since Britten and Brown's 

 Illustrated Flora of the United States and Canada is in gen- 



