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



to be used in a key except in rare instances, but it serves as supple- 

 mentary evidence in the descriptions. 



Although the texture of the leaf blades may vary somewhat with 

 the habitat and age, this is not enough to invalidate the use of the 

 character in many instances. Several groups of species are character- 

 ized by leaf thicknesses never acquired by others, no matter what the 

 growing conditions might be. 



Leaf size is too variable to be of critical value unless the discrep- 

 ancies are very great and maintained throughout several specimens. 

 Overlapping leaf sizes need not necessarily be regarded as breaking 

 down the difference if the overlapping individuals are few in number. 



Leaf shape is in many cases a trustworthy character, although in 

 some species it counts for nothing. A difference in leaf shape unsup- 

 ported by other characters is scarcely worth consideration. This 

 character alone has been responsible for more synonyms than all 

 others combined. Often, however, a perfectly legitimate species is 

 based on leaf shape supported by several minor or almost intangible 

 differences that scarcely lend themselves to description. Such in- 

 stances must be treated solely by judgment after examining all avail- 

 able material. 



The elements of leaf shape that often are of value as supplementary 

 evidence are whether (1) the leaf is broadest above the middle or 

 below, (2) the apex is acute or obtuse, (3) the base is cuneate, rounded, 

 or cordate, (4) the margins are entire or toothed, and (5) the teeth are 

 rounded (lobate), mucronate, or aristate-tipped. The presence of 

 aristate tips on the teeth is an unfailing criterion of Erythrobalanus , 

 and rounded, completely unappendaged lobes are likewise indicative of 

 Lepidobalanus . Mucronately tipped teeth are characteristic of 

 Lepidobalanus , but occasionally (in more northerly species not of 

 Central America) these tips become quite long and subspinescent, 

 but yet not aristate. (See pi. 1.) 



The surfaces of the blade sometimes show characters of great value 

 in distinguishing both species and series. The veins and secondaries 

 are characteristically impressed on the upper side in some species or 

 raised in others ; they may be very prominent beneath or only slightly so. 

 The lower surface of the lamina may be granular-bullate, low- or 

 flat-bullate, or quite smooth. The bullate condition reaches its full 

 development only in mature leaves, and therefore must be used with 

 care. (See pi. 2.) 



Pubescence may be of great significance as regards both its quantity 

 and quality and its persistence. Characteristically tomentose lower 

 leaf surfaces sometimes become nearly or quite glabrate. Some 

 species, however, exhibit a surprising degree of constancy in this 

 character. Still other species that are usually glabrate may occasion- 

 ally retain a large share of their tomentum, especially in sheltered 

 places on the under surface of the leaves. In a few instances this 

 latter circumstance has been responsible for the creation of synonyms. 

 Pubescence, like leaf shape, is a character subject to use only with 

 careful judgment. 



The number of principal veins, though a somewhat variable char- 

 acter, sometimes offers such great discrepancies as to constitute a 

 significant character. The number of veins depends to some extent 

 upon the relative length of the leaf, but this can be responsible for 



