﻿AGAVE MACROACANTHA AND ALLIED EUAGAVES. 255 



Versuch. 6, 8, 18, 142.— Baker, Gard. Chron. n. s. 8 : 780. (1877); 

 Handbook Amaryll. 193. (1888).— Ricasoli, Bull. See. Tosc. Ort. 3 : 

 306. (1878).— Hemsley, Biol. Centr.-Amer. 3: 350. (1882-6).— 

 Terracciano, Primo Contributo. 47. (1885).— Segura, El Maguey. 4 ed. 

 125. (1901). 

 A. densispina Cels, Cat. 1865. Systema. 



?? A. erubescens Ellemeet, Belg. Hort. 1871 : 1 19.— Terracciano, Bull. 

 R. Ort. Bot. Palermo. 1 : 26. (1897).— Rose in Bailey, Cyclop. Amer. 

 Hort 1 : 36. (1900). 



when young, rigfdly spreading in maturity, elongate-lanceolate to oblong, 

 at length 5 X 75 cm.: spine gray- or purple-brown, somewhat flexuous, 

 strong and rather stout, 4 X 25 mm., at length openly obliquely grooved, 

 with a low median keel, the raised sides shortly decurrent: prickles from 

 orange becoming brown, 5-10 or in maturity mostly 10-20 or 25 mm. 

 apart, narrow-based, very slender, at length 3-4 mm. long, upcurved, some- 

 times with a double flexure, the yeUow cartilaginous nearly smooth margin 

 straight between them. Inflorescence some 3 m. high, paniculate above. 

 Flowers and capsules unknown. Seeds black, very dull, large, 7-8X11 

 mm.— Plates 31 /. 2-3A. 



Mexican table-land, in the vicinity of Oaxaca. 



Specimens examined :— North of Tehuacan, Endlich, April 

 1907, no. 1918. El Parian, on the Mexican Southern Rail- 

 road,' Trelease, February 1905, mature foliage and a young 

 leaf corresponding to the type description ;— a small specimen 

 from the same collection cultivated at the Missouri Botanical 

 Garden. 



Smaller seeds, 5-6x8-9 mm., collected from an old in- 

 florescence on the road between Oaxaca and Mitla (Trelease, 

 March 1903, no. 169-03) in association with A. Karwinskii, 

 to which they were supposed to belong, have developed into 

 seedlings obviously different from the latter, and belonging 

 near to if they do not actually represent A. rubescens. It is 

 possible that young plants from the Sartorius finca " El 

 Mirador," to the northeast, collected by Purpus, also repre- 

 sent this species. 



A. rubescens, as here understood, is apparently of the alli- 

 ance of A. tequilana and its relatives of the western slope, 

 the ''zapupe" of the Southeast, and the Indian species A. 

 Cantala: the marginal prickles of mature leaves are almost 

 evppt^' those of A. Jacquiniana, but the end spine is here 



