[Vol. 9 



256 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



must have been independently developed. The septum of S. 

 longirostris is unlike that of any species of Caulanthus. There 

 is here a differentiation of the middle cells. Those near the 

 margin are elongated at right angles to the replum and are not 

 tortuous. These give way toward the middle to cells that are 

 somewhat tortuous and finally elongated parallel to the margin. 

 The narrowly winged seeds indicate a development parallel to 

 the condition in C. hcterophyllus. 



WAREA AND STANLEYELLA 



Warea is a very homogeneous group of four species that occurs 

 in the extreme southeastern part of the United States. Here they 

 seem to be confined to sandy habitats. Their remarkable isola- 

 tion from related species has undoubtedly been a factor in their 

 universal retention within a single genus, since morphologically 

 they are rather similar to species of Stanleya or Stanley ella. In 

 general appearance they resemble certain species of Cappari- 

 daceae more closely than they do any other species of Cruciferae. 

 Indeed the first species of Warea to be described was described as 

 a Cleome. 



Most of the characters of the species of Warea have been con- 

 sidered primitive in other genera. They are all annuals with 

 erect, branching stems. The leaves are all entire and in one spe- 

 cies they are deeply amplexicaul. Trichomes are not known to 

 occur on any of the species. The petals vary in color from white 

 to red-purple. The pedicels are straight and horizontally spread- 

 ing although the inflorescence is definitely corymbose and does 

 not elongate greatly even in fruit. The fruit is definitely stipitate, 

 linear, and divaricate-falcate. The stigma is subentire. 



All four species possess one character that serves to distin- 

 guish them from all related genera and, so far as the author 

 knows, from all other species of Cruciferae. The pedicels in 

 Warea are deciduous from the main axis of the inflorescence 

 and fall away attached to the mature pods. Since the line of 

 abscission is not quite even with the surface of the rachis there 

 remains a small base attached to the stem. The presence of 

 these remnants after the inflorescence has become quite mature 

 gives a characteristic roughened or knobbed appearance to the 

 axis. The mechanism by which this abscission occurs has not 

 been thoroughly studied but, from sections made of the dried 



