[Vol. 9 



242 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



var. gracilipes is the most primitive form since it possesses the 

 longest stipe. On this theory T. rhomboideum would be derived 

 from it and have given rise to T. lilacinum, T. intcgrifolium, and 

 T. affinc. From what is known of the frequently limited dis- 

 tribution of primitive species in other genera, the restricted range 

 of the variety gracilipes would seem to argue that it is a primitive 

 form. If this theory is accepted it is necessary to suppose that the 

 group in question was derived from some wide-ranging primitive 

 species of Thely podium that is now extinct. 



The second hypothesis as to the origin of "Pleurophragma" is 

 that the species developed from the same plexus as did Thely- 

 podium proper. Thelypodium rhomboideum may be thought of 

 as the most primitive form. It gave rise to T. integrijolium on 

 the north and to T. affinc on the south. One specimen has been 

 seen from western Colorado (Goodding 1789) that seems inter- 

 mediate between T. rhomboideum and T. lilacinum. In this re- 

 gion, probably, T. lilacinum was developed from the other species. 

 The variety gracilipes with its long stipe would, according to the 

 second hypothesis, be considered an atavistic variation of T. 



rhomboideum. It is then not necessary to imagine an extinct. 



wide-ranging species of Thelypodium to account for the long 

 distance of the variety gracilipes from the plexus of Thelypodium 

 proper. Except for the stipe the variety gracilipes is not more 

 primitive than T. rhomboideum. Tlie present author favors the 

 second theory as being the more probable. In either case "Pleu- 

 rophragma" could scarcely be considered worthy of generic rank, 

 but under the first hypothesis it has somewhat greater phylo- 

 genetic justification. 



Thely podiopsis. — Perhaps the most perplexing of the segregate 

 genera is "Thely podiopsis" The species transferred here are 

 very similar in appearance to Thelypodium sagittatum and at 



a casual glance souk; are likely to be confused with it. There 

 are, however, certain technical differences that should not be over- 

 looked. JMost important of all is the septum. It is here a uni- 

 form membrane quite different from that of Thelypodium with 

 its differentiated middle band. Tin; stigma in " T helypodiopsis" 

 is usually distinctly lobed with the lobes extended over the pla- 

 centae; in Thelypodium it is ordinarily quite entire or rarely may 

 show a tendency for the lobes to be extended over the valves. 

 From a phylogenetic viewpoint the two groups are well differ- 

 entiated. "T helypodiopsis" seems to have had its origin in 

 northern Arizona and the species have spread both north and 



