2 MISC. PUBLICATION 4 7 7, U. S. DEPT. OF AGRICULTURE 



the entire Central American oak flora. This same problem has necessi- 

 tated an extension of the area defined as Central America to include 

 the Mexican states below the Isthmus of Tehuantepec and the 

 Colombian Andes in South America. 



Specimens have been borrowed from the following herbaria: 3 



(AA) Arnold Arboretum, Harvard University, Jamaica Plain, Mass. 



(DeP) DePauw University, Greencastle, Ind. 



(F) Field Museum of Natural History, Chicago, 111. 



(G) Gray Herbarium, Harvard University, Cambridge, Mass. 

 (Ill) University of Illinois, Urbana, 111. 



(MBG) Missouri Botanical Garden, St. Louis, Mo. 

 (Mi) University of Michigan, Ann Arbor, Mich. 



(NY) New York Botanical Garden, New York, N. Y. 

 (US) U. S. National Museum, Washington, D. C. 



(USNA) U. S. National Arboretum, Washington, D. C. 



The author wishes to make grateful acknowledgment of his in- 

 debtedness to the custodians of these herbaria for their courtesies in 

 permitting the study of their oak material over an extensive period of 

 time. He feels particularly indebted to Paul C. Standley of the 

 Field Museum and Robert E. Woodson of the Missouri Botanical 

 Garden whose numerous favors greatly facilitated the completion of 

 the work. S. F. Blake of the Bureau of Plant Industry rendered 

 valuable aid and criticism, particularly with the Latin diagnoses. 



SUBGENERIC CONCEPTS 



The subgenera outlined by Trelease seem in no need of revision, but 

 the arrangement of species within them may be improved in the light 

 of recent studies. The treatment of the subgenera as genera by 

 Schwarz (4, 5) seems to have been done without sufficient reason or 

 advantage. 



Only two subgenera occur in Central America. They are Lepidobal- 

 anus (in place of which Trelease applied Engelmann's synonymous 

 name, Leucobalanus) and Erythrobalanus . Lepidobalanus includes the 

 European Quercus robur L., the type species of the genus. Oersted's 

 subgenus Macrobalanus , which Schwarz has raised to generic rank, also 

 belongs here. It differs from Lepidobalanus only in having unequal 

 cotyledons (and consequently lateral and obliquely directed radicles) 

 and in no sense constitutes a group that one could term generically 

 distinct from Quercus, subgenus Lepidobalanus. 



There might be a more convincing argument for the generic distinc- 

 tion of Erythrobalanus, for it is distinguished from Lepidobalanus by 

 leaf, flower, cup, and acorn characters, which seem rather important. 

 However, it is quite clear that Schwarz proposed its generic distinction 

 without first familiarizing himself with all the species, for his first 

 publication on the subject referred the Colombian oaks to Lepidoba- 

 lanus. The following year he concluded that they belonged in 

 Erythrobalanus. Also in his first publication he referred the series 

 Durifoliae, Scytophyllae, Costaricenses, Citrijoliae, and Borucasanae to 

 Lepidobalanus. These series are composed of species that, for the 

 most part, are very definitely members of Erythrobalanus. He seems 

 to have placed them in Lepidobalanus solely upon the basis of their 

 lateral abortive ovules and without regard for the several other more 



3 The letters in parentheses are used in the citation of specimens and the explanations of the illustrations 

 to indicate the herbaria from which material has been available for study. 



