APR. 1960 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 10(2):58-64 



TAXONOMIC RELATIONSHIPS OF DIOMMA ENGLER EX HARMS 



Richard S. Cowan and George K. Brizk ky 



Department of Botany, U. S. National Museum, 

 Smithsonian Institution, Washington, D. C. 

 and 



Yale University School of Forestry, New Haven, Connecticut 



INTRODUCTION 



The formal taxonomie conclusions reached in this study can hardly be pre- 

 sented without acknowledging the great debt owed Stern and Brizicky for the 

 very lucid treatment of the anatomy and morphology of this group and its rela- 

 tives reported in a preceding- paper of this serial.* Much of the history as well 

 as the morphology and anatomy of the groups involved is quite adequately set 

 forth in their contribution and will not be repeated here. However, since three 

 former genera are herein reduced to subgenera of a single genus, the reasons for 

 such a modification should be summarized. 



The key to the subgenera contrasts the only known differences of any magni- 

 tude; it is at once apparent that these are few and in most instances more or less 

 continuous, rather than discontinuous as one may justifiably expect differences of 

 generic stature to be. The many similarities among the three taxa are amply sum- 

 arized by Stern and Brizicky and it appears obvious that the similarities far 

 outweigh the dissimilarities. The decision that incipient rather than completely 

 former genera are herein reduced to subgenera of a single genus, the reasons for 

 such a modification should be summarized. 



The question may be raised that, since this is obviously a monophyletic group, 

 we must account for the present geographic distribution of its parts. If we may 

 place some reliance on the maps presented by Schuchert (1935) showing the 

 configurations of the land masses in and adjoining the Antillean — Carribean 

 region, more or less continuous land masses have been available for migration (via 

 Central America) between Cuba and the Bahamas at one extreme and northern 

 South America on the other for long periods, beginning in the Late Upper 

 Cretaceous and extending w T ell up into the Eocene. Simpson (1940) considers 

 the land connection between North and South America to be a much more recent 

 event — perhaps as late as the Pliocene — but this conclusion does not invalidate 

 the supposition that migration was by a land route between Cuba and northern 

 South America, at least in part. Indeed, the fact that the majority of Cuban 

 species have their closest affinities in South America (Alain 1958) clearly indi- 

 cates rather efficient intercommunication between the two extremes. In addition 

 to migration over land routes, the powerful winds of Caribbean hurricanes must 

 have played an important role in the dispersal of plants in the West Indian — 

 northern South American region. The winged fruits of all species of Spathelia, 

 for example, are susceptible to wind-dispersal, and the present distribution of the 

 genus may be the result of the operation of this agent alone. Whatever mecha- 

 nism was responsible, it is certainly less novel than the direction of the migration, 

 which is unusual but characteristic of Simpson's "filter land-bridges" (1940), 

 through which migration always occurs in both directions. 



* 10: 38-57. 



-58- 



