84 



cens, with some Q. Fendleri. At the lower levels Q. Rydbcrgiana 

 and Q. grisea were also present. Of course the explanation of 

 this distribution is found in the fact that the oaks of the Gambelii 

 series belong to a colder climate and ordinarily to a higher alti- 

 tude than those allied to undulata. On May \J Q. nitescens and 

 Q. Gambelii were practically over flowering, and Q. Novomexi- 

 cana was coming into flower ; while the undidata series (i. e., all 

 the other species) were in bud or little more advanced. I will 

 now consider the species somewhat more in detail : 



(A.) Gambelii Series 

 In the early leaf, the three forms are easily separated thus : 



Lobes of leaf not at all bifid ; general color light green with little red. Q. Gambelii. 

 Lobes of leaf (except towards base) bifid. 



Young leaves very light green, narrower, deeply incised, hanging down. 



Q. Novontexicana. 

 Young leaves not so light, broader, not so deeply incised, not so pendulous. 



Q. nitescens. 



Q. nitescens has the young shoots dark red, and the midribs of 

 the leaves usually reddish, but the leaves glossy light green. 

 Q. Novomexicana has the midribs green ; the young leaves are 

 quite grayish compared with the other two forms. Q. Gambelii 

 has narrow pendulous leaves, but is easily distinguished from Q. 

 Novomexicana by the characters already given. Altogether, the 

 new evidence tends to substantiate the validity of these three 

 species, about which I previously felt doubtful. Dr. Rydberg 

 says the leaves of Q. Novomexicana are bright red when they 

 unfold, but I am sure this could not have been the case with the 

 plants I studied. I am strongly convinced that the pigmentation 

 of oak-leaves varies to a great extend independently of the other 

 characters, as DeVries so often found with the pigmentation of 

 flowers.* Under these circumstances, the name must go with 

 the leaf-form rather than with leaf-color. 



(B.) Undulata Series. 

 Qucrcus undulata Torr. Leaves still very small, but as in 

 all of this series except Q. Fendleri, some leaves of last year 



* Cf. his statement : " The units of the specific characters are to be regarded and 

 studied as shanply separated quantities." (Jour. Roy. Hort. Soc. 25: 243. 1901.) 



