1922] 



PAYSON — STUDY OF THELYPODIUM AND ITS IMMEDIATE ALLIES 257 



material, seems to be due to a simple breaking apart of the cells 

 in a certain region. 



The septum in Warea is somewhat similar to that seen in 

 Thelypodium. There is a differentiation between the cells near 

 the margin and those in the middle. The latter are somewhat 

 tortuous and considerably elongated. The walls are not densely 

 compacted as in some species of Thelypodium. 



be considered most primitive 



of the am 



Florida. It 



& th of ,uu, „x fe 



does not, however, possess the longest stipe, although there is 

 no considerable difference between the various species in the 



_ tn. It is scarcely possible to trace any devel- 

 opmental sequence between the species that may be correlated 

 with geographical distribution. This is of course partly due to 

 the limited variation between the species and to their restricted 

 range. Warea Carteri may well be an offshoot from W. cuneifolia. 



Stanleyella is evidently most closely related to Warea but may 

 be considered distinct from that genus because of the persistent 

 pedicels, the much shorter stipe, and the different septum. The 

 septum in Stanleyella is somewhat differentiated between margin 

 and middle, but the cells are not tortuous as in Warea. The 

 leaves in Stanleyella are toothed, while in Warea they are entire. 

 Trichomes are present in some individuals of Stanleyella but 

 have not been noticed in Warea. Stanleyella may be restricted 

 to the southwestern and Warea to the southeastern portions of 

 the United States. Stanleyella is undoubtedly to be considered 

 less primitive than Warea, but because the absciss mechanism in 

 the pedicels is lacking it was probably developed from a com- 

 mon ancestor and not from anv of the present species of Warea. 



Key to the Genera 



The genera treated in the present paper may be contrasted 

 and their chief morphological characters summarized by the 

 following key. This key, it must be understood, makes no at- 

 tempt to account for genera that are closely related but have not 

 vet been elaborated. 



A. Sepals spreading or reflexed in anthems; glabrous or very 



sparsely pubescent, annual or biennial herbs; blade of 

 petal entire; Btigma entire or nearly so; pedicels slen- 

 der, divaricate; pods widely spreading. 

 a. Stipes ~> nun. or more long; leaves entire; pedicels 



becoming detached from the raehis at maturity WAREA 



