4 MISC. PUBLICATION 17 7 



which he referred a single species are in most cases scarcely to be held 

 separate from some neighboring series. The series Cyclobalanoideae, 

 for instance, is not only synonymous with the Corrugatae, but its one 

 species, Quercus cyclobalanoides, is synonymous with Q. corrugata. 



In a consideration of the Central American species alone, no suit- 

 able line has been discovered along which the species could be divided 

 into sections intermediate in rank between the subgenera and the 

 series. 



It is apparent that the morphological basis of the classification of 

 Quercus offers too little evidence of the best lines along which the 

 many species may be segregated into greater groups. The distinction 

 between the typical subgenus Lepidobalanus and the subgenus Eryth- 

 robalanus is not wholly satisfactory. That between the various series, 

 even though a real effort has been made to include in them natural 

 groups, is even less so. 



SPECIES CONCEPT 



To discuss one's species concept without defining "species" perhaps 

 seems a bit negligent. Considering the facts that very few taxon- 

 omists agree on the definition of the term and that the phenomenon 

 (if it is a real one) varies so much from genus to genus as to necessitate 

 drastic changes in concept from group to group, it seems useless to 

 attempt more than a statement of position in general terms. 



Species in Quercus seldom exhibit the same degree of distinctness 

 from one another. For instance, Q. adata is not closely related to any 

 other known species, and Q. tristis is only doubtfully distinct from 

 Q. castanea. There is therefore no sharp line (or no line at all) on 

 one side of which the degree of distinctness constitutes varietal rank 

 and on the other specific rank. It is with misgivings that the varietal 

 category is used in this treatment at all. Only the basis of the forma, 

 which is not employed in this publication, can be clearly stated, and 

 that merely because it embodies a phenotype of simple genetic differ- 

 ences. 



Even within a single genus it is sometimes necessary to admit more 

 than one concept in defining species. This is especially true in Quercus, 

 which contains widely ranging, ecologically tolerant species and often 

 closely related ecologically intolerant species of restricted range. In 

 either case the species must be defined so as to allow for the maximum 

 of variation compatible with a sense of order. In the Atlantic region 

 of the United States occur several species of the first class that, be- 

 cause of their wide ranges and consequent minor differentiations, 

 have been described repeatedly with no better result than the multi- 

 plication of synonyms. In the United States-Mexican border region, 

 on the other hand, there exists a condition induced by the mountainous 

 terrain and characteristic basin and range topography that has favored 

 a relatively high percentage of endemics. In some cases quite distinct 

 species are known that exhibit less profound differences than the 

 variation often encountered within a single species of the Atlantic 

 United States. Of the border region species, the more xeric are usually 

 the more widespread, since the less xeric types (which are often en- 

 demics) are confined to the relatively scarce moist, or at least mesic, 

 habitats. 



