(i88) 



ground the colonies are more irregular and shrub-like. 

 When the leaves unfold in the spring the contrast between 

 the clumps of the different species or subspecies, as you 

 may call them, is best shown, as one clump is bright red, 

 while the next may be yellowish or gray. One of the yellow 

 forms I afterwards identified as J^. Gambclii Nutt. and the 

 gray one is J^. Fcndlcri Liebm. 



In the South Cheyenne Canon, near Colorado Springs, in 

 Wahatoga Canon between the two Spanish peaks and on 

 Turkey Creek, a tributary of Huerfano River, we also found 

 oaks but of a different habit, viz., solitary, middle-sized trees, 

 that could scarcely be referred to J^. Gambclii. Those of 

 the last locality, at least, were so unlike any of the forms in- 

 cluded in J^\ Gambclii, that I immediately regarded them as 

 belonging to an undescribed species ; the thin leaves and 

 general habit suggest rather J£. minor. 



When I left I instructed Mr. Vreeland to collect acorns in 

 the fall. He has sent me specimens of three or perhaps four 

 distinct forms of the J^. Gambclii series. We may call them 

 species or varieties just as we please, but they look very dif- 

 ferent in the field — more so than in herbarium specimens. Be- 

 sides these I had notes and material gathered by myself in 

 1895, by Dr. Fred. Clements, of Nebraska, and Mr. E. 

 A. Bessey, of Washington, D. C, in the years 1896-1900, 

 and of course the specimens in the herbaria of New York 

 Botanical Garden and Columbia University. I found, how- 

 ever, that these collections were not sufficient to make a 

 thorough study of the oaks of the southern Rockies. I wrote 

 to the Missouri Botanical Garden and the United States Na- 

 tional Herbarium, asking for the loan of their specimens of 

 J^. undulata, ££. Gambclii and related species. These re- 

 quests were kindly granted. 



My original intention was to confine my study to the oaks 

 of the Rockies proper, but I found that in order to reach a 

 satisfactory understanding of these, it was necessary to take 

 into consideration the nearly related forms of the plains and 

 table-lands of Arizona, New Mexico and western Texas. I 



