1922J 



PAYSON — STUDY OF THELYPODIUM AND ITS IMMEDIATE ALLIES 239 



Leaves. — The species of Thelypodium may be grouped in three 

 divisions according to the shape of the stem-leaves. In nine 

 species they are entire and amplexicaul at the base. In T. lacinia- 

 tum and its varieties they are irregularly toothed or lobed and 

 more or less petioled. In the remaining four species the leaves 

 are entire but not auriculate at the base. Within the genus there 

 is in all the known species a differentiation between cauline and 

 radical leaves. The radical leaves are generally oblanceolate in 

 outline and entire or subentire. In only four species are they 

 conspicuously toothed, namely, T. brachycarpum, T. crispum, 

 T. Howellii, and T. laciniatum. In many species a definite petiole 



is present. 



Among those species having entire amplexicaul stem-leaves the 



most primitive in respect to the length of the stipe, the type of 

 inflorescence, and the form and color of the petals is undoubtedly 

 T. eucosum; the most specialized in these and other characters 

 is T. flexuosum. In the former the basal lobes are well developed, 

 in the latter they are more reduced than in any other species of 

 this group. From T. eucosum the other species with amplexicaul 

 stem-leaves may be derived without difficulty. Among the spe- 

 cies having entire but not amplexicaul stem-leaves it would be 

 hard to say which is the most primitive. That there is no great 

 gap between this group and the first is shown by the fact that in 

 T. lUacinum (and probably in others) small auricles are some- 

 times present at the base of the leaves. It seems more probable 

 that this type of leaf was derived by the reduction of the basal 

 lobes than that the change was in the other direction. It remains 



now to consider the third type of leaf— the lobed and petioled 

 stem-leaves of T. laciniatum. Unlike T. eucosum, T. laciniatum 

 has no near relatives. It scorns to represent an ancient offshoot 

 from primitive stock rather than to be the ancestor of species now 

 extant. It may have been derived from a species with amplexicaul 

 leaves but there is no reason to suppose that it has given rise to 

 species with that type of leaf, since they can be traced to a differ- 

 ent ancestry. The entire amplexicaul stem-leaf is considered the 



om it the others are believed 



rived 



h 



of stem-leaf were primitive ; it is not surprising to find 



same order of develc 



same 



Because of comparative uniformity in the differentiated radical 

 leaves in this genus there is little evidence relating to the steps 



