182 



MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



[VOL. 10 



it appears to have a distribution rather different from that of the Mutisieae con- 

 sidered. For example, in Carduus nutans (Daniel), sclerenchyma is on the abaxial 

 face of the bract, whereas it is found most predominantly on the adaxial side 

 in Mutisieae. In Heliantheae subtribe Verbesininae, strands of thin-wall fibers 

 may be present (Napp-Zinn). Although fibrous bundle sheaths are not present 

 in bracts of the helianthoid Fitchia (Carlquist 1957d), sclerification of the 

 inner face of the bract and development of diffuse sclereids are not unlike what 

 is found in certain Mutiseae, providing- the closest parallel available from other 

 Compositae. 



SUMMARY 



The data of this study represent several new anatomical records for Mutisieae 

 (laticifers, vascularized receptacular bracts) as well as providing new descrip- 

 tions of the anatomy of genera in the tribe. The facts developed in each category 

 — stem, node, leaf, and involucre — definitely support the distinctness of the 

 genera recognized by Maguire et al. (1957). Leaf anatomy, for example, could 

 be used for the identification of most of these genera. In many instances, excel- 

 lent specific characters are provided, e.g., characters of leaf anatomy in Steno- 

 padus, Stomatochaeta, and Duidaea. Because of the seemingly greater variation 

 in vegetative and involucral anatomy than that found in flower and pollen 

 structure (Carlquist 1957d), many of the characters developed in this study 

 tend to be of greater importance at the specific level. 



In regard to broader relationships, structural details confirm the close rela- 

 tionship of Stenopadus with Stomatochaeta and Chimantaea, as well as the 

 similar affinities of Achnopogon, Duidaea, and Neblinaea with Gongylolepis. In 

 particular, the presence and distribution of sclerenchyma in stems and involucral 

 bracts as well as the presence or absence of laticiferous cells are significant in 

 uniting the genera within each group, as well as in differentiating the two 

 groups from each other. 



The presence of laticiferous cells in Quelchia is rather curious, considering 

 that these are characteristic of Gongylolepis and its allies (Gerberinae) rather 

 than the subtribe to which Quelchia is assigned, Gochnatinae, and it may be 

 closer to Gerberinae than other genera of Gochnatinae. The characters of 

 (ilossarion show that it is obviously related to the remaining genera of Guayana 

 Mutisieae, but not to any particular one of them. Its differentiation appears 

 to have b«en separate from either the Stenopadus or the Gongylolepis lines, a 

 conclusion also reached by Maguire (1956) on the basis of his studies. 



Reference is made in this study to other genera of Gochnatinae — Anastra- 

 pliia, Hesperomannia, Stifftia, and Wunderlichia — in an attempt to show the 

 relative uniformity of the Guaj r ana group and also to find similar characters 

 in non-Guayana genera. Although Stifftia seems advanced in its involucral bract 

 anatomy, its leaves are unspecialized. The sum of its characters suggests that 

 it belongs among Gochnatinae near Stenopadus. The wood of Stifftia (Carlquist 

 1957b) is quite similar to that of Hesperomannia, which has leaf anatomy like 

 that of Stifftia. In characters of involucral bracts and the non-curling habit of 

 corolla lobes, Hesperomannia and Anastraphia agree. Anastraphia, however, has 

 specialized leaf anatomy. The stem anatomy of Hesperomannia is much like that 

 of Stenopadus. Hesperomannia, then, is exemplary of the complex relationships 

 within this group of primitive Gochnatinae. Likewise, Wunderlichia has re- 

 tained a primitive floral venation while becoming specialized in its advanced 



