34 THE CACTACEAE. 



ish, the upper generally weaker, the 2 or 3 uppermost ones much weaker, clustered closely together 

 and very pale, some of them sometimes crowded towards the center, the central spine solitary, 

 very stout and erect or sometimes curved or even hooked, yellow, 3.5 cm. long; all the larger spines 

 somewhat bulbous at base ; flowers 5 to 6 cm. long, salmon-colored ; ovary 20 to 25 mm. long, bearing 

 4 to 7 minute caducous scales ; fruit narrowly oblong, 6 cm. long ; seeds large, 3 mm. long, shining. 



Type locality: Cited as Sonora in first publication of species; afterwards as south side 

 of the Baboquivari Mountains in northern Sonora. 



Distribution: Mountains of southern Arizona, southwestern New Mexico, and northern 

 Sonora. 



We have followed Mrs. K. Brandegee in referring Alammillaria hrownii here, for not 

 only do the original descriptions read much alike but the type localities for the two are in 

 the same mountain range. M. hrownii was described from a very small plant and dif- 

 fers considerably from mature individuals. Engelmann calls attention to the very large 

 seeds, which he says are "larger than those of any other Mammillaria examined." He 

 also states, "embryo with some albumen, curved; cotyledon foliaceous, approaching the 

 structure of the seed of most Echinocacti." 



Dr. Shreve reports that the flowers appear in the summer and the fruits, which follow, 

 hold over the following winter, gradually drying up. The fruits do not open by a basal 

 pore as in other related species. 



We would refer here specimens from Lordsburg, New Mexico, and Bowie, Arizona, which, 

 have heretofore been referred to Mammillaria valida, now Coryphantha muehlenpfordtii. 



Illustrations: Bot. Gaz. 22: 254, as Mammillaria hrownii; Cact. Journ. i : 85 ; Cact. 

 Mex. Bound, pi. 74, f. 8, as Mammillaria rohustispina. 



16. Coryphantha connivens sp. nov. 



Globular or somewhat depressed, 8 to 10 cm. broad, somewhat woollj^ at the crown at flowering 

 time but becoming glabrate; spines all radial but of two kinds; one kind 5 or 6, spreading or cur^^ed 

 backward, subulate, horn-colored, the other 8 to 10, from upper part of spine-areole, clustered, erect, 

 or toward top connivent, acicular, black at tip; flowers yellow, 6 to 7 cm. broad; perianth-segments 

 narrowly oblong, acuminate; fruit greenish, oblong, 3 cm. long; seeds brown, oblong, 2 mm. long. 



This species is common in the Valley of Mexico, especially on the pedregal. Dr. Rose 

 collected it first in 1901 and again in 1905 and 1906; the type is his No. 8372 from near 

 Tlalpam, collected in 1905. Dr. C. Reiche also collected it between Tacubaya and 

 Santa Fe in 1922, and according to him the plant from this locality is the one referred to 

 Mammillaria pycnacantha by Schumann (Gesamtb. Kakteen 489. 1898). 



The species is characterized by the peculiar clusters of spines in the upper angle of 

 the areoles. A small plant was sent by O. Solis from Tlalpam in 1907, but it has fewer 

 acicular spines than described above. 



17. Corjrphantha pectinata (Engelmann). 



Mammillaria peclinata Engelmann, Proc. Amer. Acad. 3: 266. 1856. 

 Mammillaria pectinata cristata Hortus in Forster, Handb. Cact. ed. 2. 403. 1S85. 

 Cactus pectinatus Kuntze, Rev. Gen. PL i: 259. 1891. 



Usually simple, globose, 3 to 6 cm. in diameter; tubercles usually arranged in 13 spirals; upper 

 tubercles 10 to 12 mm. long, about twice as long as lower ones; areoles a little longer than broad; 

 spines 16 to 24, all radial, those on lower areoles appressed and often a little recurved, those from 

 upper part of upper areoles 12 to 18 mm. long, connivent over apex, yellowish white with black 

 tips; flowers yellow, 5 cm. long; ovary 6 to 8 mm. long; fruit 12 mm. long. 



Type locality: On the Pecos River in western Texas. 



Distribution: Southern Texas and adjacent parts of Mexico. 



Coulter and Schumann refer it to Mammillaria radians De Candolle, but it doubtless is 

 a distinct species. 



This plant is well illustrated by Engelmann and should be easily recognized. It 

 appears to have been collected only rarely. The only representatives we have of 



