BOTANICAL GAZETTE. 236 



respects it represents the typical form of E. Cahfoniica, but the per 

 ennial rootstock seems to distinguish it ; annuals, to be sure, in mild 

 climates not rarely last for several years, e. g. Sn/aiiu/u i/i):;nii/i in 

 soutliern California, but in these tiie stem becomes ligneous and no 

 rootstock forms, the normal tap-root not losing its characters, even if 

 it does become 3 "r 4 years old. It is barely i)ossible that E. Cali- 

 foniica is one of tiiose perennials which will llower as seedlings in the 

 first years and that then the aridity of the climate in many instances 

 kills it, root and all; but if so, why has this occasionally (?) perennial 

 character not been observed before? 



PoRTULACA SUFFRUTKSCENS, n. sp.— Erect, about a span high 

 from a stout, branching and ajiparently ])erennial rootstock, ligneous 

 at base; leaves terete, about i inch long, with sparingly hairy axils; 

 flowers clustered at the end of the branches, large (7 to 10 lines wide), 

 yelloAish buff colored; petals obcordate or em.irginate ; stamens num- 

 erous; filaments, like the 5 or 6 stigmas, red; seeds dark, with me- 

 tallic luster, tuberculate. 



In western New Mexico, at the copper mines, C. JVriij^/it, 874, 

 coll. 1851 ; Cones and Paliiicr, Fort \Vhi|)ple, northern Arizona, 1865; 

 found by myself 1880, on rocky banks in the Santa Rita Mountains, 

 southern Arizona. 



Very near P. pihsa, with which I had formerly united it ; the 

 seeds of both are similar, their tubercles, magnified 40-60 diameters, 

 ai>pear veiy prettily as overlapping excrescences with a toothed free 

 edge ; both have dark seeds, ours with a metallic luster, the 

 others more dull. The number of stamens in different (lowers was 

 about 40 ; while in P. pilosa it is stated to be 15 to 25, but in cultivated 

 s|)ecimens of the latter I ha\e found as many as 50! The color and 

 .size of the flower, the larger leaves with sparing hair in the axils, and 

 the stouter stems and perennial (?) rootstock distinguish it from its pur- 

 ple flowered annual relative. 



Rosa simthamea, Watson, Fl. Cab, 2, 444. — In the deep shade 

 of the l>i'f Trees oi Fresno county, Cal., where scarcely anything else 

 grows. I found what I take to be a form of this: pretty little species, 

 blooming in September. I may designate it as 



Var. suniNiiRMis: Stems a span high, glabrous or more or less 

 glandular hispid, with a few scattered setaceous spines (none stipular) 

 or spineless ; stipules short and narrow with short narrcnv divaricate free 

 j)c)ints; leaflets mostly 5, thin, pale below, elliptical or nearly orbicular, 

 obtuse, sharplv serrate and glandular serrulate; rhachis glandular-pu- 

 bescent and often sjMnulose, petiolule of terminal leaflet almost its own 

 length ; flowers single {\){ to \)A inches wide) rose colored ; i)eduncle 

 slightly glandular, calyx tube globose, naked, lobes entire. 



The stems of the same season bear the flowers, as is the case in 

 R. foliolosa of Texas; or is it in this instance an autumnal form ? There 

 may be characters enough, especially the absence of any stipular 

 spines, to distinguish specifically this southern form from the northern 

 type, but considering the great variability of roses it is thought best to 

 keep them together for the present. 



