1922] OA 1 



PAYSON — STUDY OF THELYPODIUM AND ITS IMMEDIATE ALLIES ^41 



included in it. No attempt will be 

 to nlace all these species generically 



All that now may be said 



many of them is that they may not be included 



diuvi. This does 



&ic ft o,i^ 6 



proposed by Dr. P. A. Rydberg will necessarily be main- 

 thft nrosent author — it does mean that their affinities 



tained by the present author — it 

 are not with this genus. 



Pleurophragma. — This group was proposed as a genus by Dr. 

 Rydberg to include Thelypodium integrif olium and its immediate 

 allies and was based primarily on the presence of a "strong and 

 broad midrib of the septum of the pod." It was also stated 

 that "there is no distinct midvein in any of the typical Thely- 

 podia." Another character that might be used to support Pleuro- 

 phragma is found in the leaves. They are here sessile, entire, 

 and not amplexicaul — a combination that obtains 

 of Tlielvv odium except in this group. That this cl 



while 



f 



shown by the occasional appearance of auriculate lobes at the 

 base of the leaves in individuals of T. lilacinum. In these ex- 

 ceptional cases the leaves are not unlike those of the amplexicaul 

 group. The value of the "strong and broad midrib of the sep- 

 tum" is scarcely of more importance. The cellular pattern of 

 the septum has already been discussed, and it has been shown 

 that in ail the species of Thelypodium there is a differentiated 

 middle region where the cells are elongated parallel to the replum 

 and the cell-walls more or less closely compacted. No other 

 "pattern" occurs in the species transferred to Pleurophragma al- 

 though here the extreme in differentiation is probably reached. 

 The "strong and broad midrib" results from the fact that in these 

 species the dense middle zone is colored yellow and so stands 

 out in contrast to the marginal cells. Furthermore, this differ- 



ential coloring of the middle region is not peculiar to T. integri- 

 f olium and its allies. It is also seen in T. flcxuosum and in indi- 



T. crispum. From 



distinct 



it would seem impossible to retain Pleurophragma a 



from Thelypodium. 



The phylogenetic relationships of the species of " Pleurophrag- 

 ma" among themselves are not clear. This is due in part to the 

 slight differentiation between the species and in part to the evi- 

 dence offered by T. rhomboideum var. gracilipes. Two 



im var. gracilipes. Two hypothe- 

 possible: first, it might be supposed that the group 

 inated in western Colorado or Utah and that T. rhomboideum 



seem 



