114 PROCEEDINGS OF THE ACADEMY OF [Jan., 



woodsii, but I am convinced that the value of this character has been 

 much exaggerated. Crepin in liis table (1896) indicates the variabilitj^ 

 of pisocarpa in this respect. 



(2) E. arkansana Porter. 



Much emphasis has been placed upon the statement in the original 

 description that the sepals are reflexed in fruit. This seems to dis- 

 tinguish my plant from the true arkansana (the fruit was miknown 

 whemi consulted Dr. Rydberg), but I strongly suspect that, as is usual 

 in the group, the sepals are more or less reflexed at first, but ultimately 

 erect. Another discrepancy is in the leaflets, which are 9 to 11 in true 

 arkansana, but 7 to 9 in my plant. The other characters seem very 

 similar in both plants. 



(3) R. fendleri Crepin. 



I regret very much that I cannot make sure of fendleri, but this is 

 perhaps not surprising, as Crepin himself did not know what to do with 

 it twenty years after its description. Dr. Rydberg writes: ''My idea 

 ofJRJfendleri is a low shrub with evident infrastipular spines, and with 

 stipules, petioles and peduncles glandular." No such rose as this was 

 founcljat' Pecos. Rehder's account of R. fendleri (in Bailey's Cyclo- 

 pedia of American Horiicidture) indicates a totally different plant from 

 that indicated by Dr. Rydberg, much more in the manner of R. peco- 

 sensis. However, the sepals are said to be quite entire, and the flowers 

 normally in clusters. 



(4) R.'^gymnocarpa Nuttall. 



Easily separated from ours by the deciduous sepals. 



(5) R. neomexicana Ckll. 

 Fruits oblong; sec above. 



(6) R. aciculata (Ckll.). 



Wet Moimtain Valley, Colorado. Described as R. blanda var. 

 aciculata in Science Gossip, 1SS9, p. 18S. Type at Kew; photographs 

 distributed to several herbaria. This has the flowers not very large, 

 bright pink; prickles straight, slender, pale; sepals entire, hispid; 

 stipules dilated, entire or slightly toothed at apex; leaflets 5 to 7, broad- 

 oval, simply'serrate, pale beneath. There are no paired infrastipular 

 prickles. This appears to bo allied to R. sayi, which occurs in the same 

 vicinity, but the leaves arc very much smaller, and their marginal 

 teeth are normally simple. 



(7) R. sayi Schwein. 



Low, with large, thin, doubly-serrate leaflets; grows in moist places, 

 and occurs as far south as Beulah, New Mexico. Just how far the char- 



