EXGELMAN.V — OAKS OF THE UNITED STATES. 373 



or retuse, often again 2-3-lobed. They bear middle sized or small 

 oval acorns in more or less knobby hemispherical cups. Scattered 

 copses of these broad-leaved oaks, often of a beautiful brownish- 

 purple in September, accompany us to within a few hundred 

 yards of the top of the canon, but here the character of these shrubs 

 changes : the bushes are lower, the leaves smaller and in outline 

 narrower, the lobes narrower and mostly undivided, but still ob- 

 tuse. Now we near the precipice itself; from the ragged, dizzy 

 edge we here and there get a glimpse of the young Arkansas, 

 whose clear green waters toss and foam, twelve or fifteen hundred 

 feet under us, through the inaccessible gorge, rushing towards the 

 plains. The oak bushes accompany us even here, but now they 

 are only 4-6 feet high, with leaves 2 inches long, ovate-lanceolate 

 in outline, no longer lobed, but coarsely dentate, the acute teeth 

 terminating in a sharp point ; the acorns are scarcely different 

 from those noticed before. A few steps more and we have 

 reached the brink of the precipice itself: oak bushes here too, 

 but only 3 or 4 feet high, with small (i inch long), oval, firm, 

 almost cartilaginous, semipersistent, spiny-toothed leaves, here 

 and there with only very few teeth or quite entire ; the acorns 

 proportionately smaller, of the same short oval shape, or often 

 elongated from an unusually small, scarcely knobby, and some- 

 times peduncled cup. 



We feel satisfied that we might have abundant material to 

 characterize several distinct species, certainly 4 or 5 well marked 

 forms, and, indeed, they have been considered such. The first is 

 Nuttall's ^uercus Gambelii (^. stellata, var. Utahensis^ D. C, 

 Prod.) ; the second is ^. alba^ var. Gunnisoni of Torrev ; the 

 third, with acutish lobes or coarse teeth, is Torrey's old ^. un- 

 dulata of Long's Expedition, the first oak obtained from these 

 mountains, and described about fifty years ago ; the fourth, from 

 the edge of the precipice itself, is what hjis often been mistaken 

 for Torrey's Q. Emoryi^ or what has been named ^. pungens^ 

 Liebm., in part ; with it occur entire-leaved forms which seem 

 to unite with this as a fifth form the Q. oblongifolta^ of the same 

 author, and ^. grisea^ Liebm. As a large and broad-leaved 

 southeastern form somewhat allied to ^. Gambelii I consider 

 J^. Druminondii, Liebm. In herbarium specimens they all 

 appear distinct enough, but, looking around us, the very abun- 



