1957] 



ANATOMY OF GUAYANA MUTISIEAE 



475 



no suggestions have been made previously concerning the primitive carpellary 

 structure in Compositae. One might expect that the ancestral condition for Com- 

 positae would have an interior set of bundles, representing carpellary vascular 

 supply, as well as an exterior set representing that of perianth parts, as well as 

 of the stamens. As the descriptions in the main body of this paper indicate, this 

 is indeed the case in the majority of these Guayana Mutisieae. Furthermore, it is 

 interesting that many of these genera show a characteristic grouping of these 

 bundles at the base of the achene. The dorsiventral pair — representing the main 

 stylar traces, and homologous to median traces of the carpels supposed to have 

 been primitively present in Compositae by Eichler (1875) — often join exterior 

 bundles, which form a plexus at the base of the ovule trace. The large number of 

 lateral interior bundles in some of the taxa considered here (Stenopadus kun- 

 bardtii, for example) may represent a secondary increase in the number of such 

 bundles, but the frequent presence of a lateral pair as well as a dorsiventral pair 

 suggests that at least these four may have been present primitively. 



A functional correlation might explain the fact that lateral interior bundles 

 often closely accompany the two bands of stigmatoid tissue on the interior of the 

 achene wall. The double nature of these lateral interior bundles beside stigma- 

 toid tissue in several of the taxa described here (see Stenopadus cucullatus, for 

 example) may be a further suggestion of a two-carpellate condition. Some lateral 

 bundles often develop phloem only, suggesting (as in filaments of many Compo- 

 sitae) that loss of xylem precedes loss of phloem in the progressive reduction of 

 vascular tissue. 



In the Heliantheae, the writer (Carlquist 1957b) proposed that the presence of 

 four bundles in the style represents a primitive condition. The connection of the 

 style bundles to interior, or carpellary, bundles in the achenes of Mutisieae 

 would seem to substantiate this suggestion. Differential loss of vascular tissue 

 in different parts of the flower could explain why lateral carpellary bundles may 

 be present in some of the taxa studied here {Stenopadus, for example) without 

 lateral style bundles. Stenopadus cucullatus would represent a species in which 

 this reduction has not occurred. 



Also on the basis of study of primitive Heliantheae, the writer considered that 

 a dichotomous ovule trace might be a primitive character, and that ancestrally two 

 ovules may have been present. The frequent occurrence of a dichotomous ovule 

 trace in the Mutisieae studied here would seem to offer an additional instance of 

 this condition. The elaborately branched ovule trace of Stenopadus kunhardtii, 

 however, is probably best interpreted as a secondary increase in the vasculariza- 

 tion of the ovule. 



The vascular patterns described above would seem to require an explanation 

 in evolutionary terms, and other studies are desirable in supplying a comprehen- 

 sive picture of evolution in the primitive Compositae. The writer wishes to em- 

 phasize the agreement between data from gross morphology assembled by Maguire 

 and his colleagues and the anatomical descriptions given here. Further studies, 

 particularly on stem and wood anatomy, are planned to aid in the development of 

 a thorough understanding of this group. 



Claremont Graduate School 



Rancho Santa Ana Botanic Garden 

 Claremont, California 



