﻿AGAVE MACROACANTHA AND ALLIED EUAGAVES. 251 



between them. Inflorescence about 3 m. high, laxly oblong-paniculate 

 in the upper half: pedicels 2 X 1-3 nun. Flowers very glaucous, green, 

 with a slight purple tinge: ovary 6-7X 20-25 mm.: tube of penanth 

 oblong-conical, 8X 12-15 mm., bearing the filaments about its middle; 

 segments mostly arcuately incurved, 3 X 15 mm.: style and filaments 

 abundantly dotted with purple, 30-40 mm. long: anthers yellow-green 

 purple-dotted. Capsules blue glaucous, broadly oblong, attenuate at each 

 end, stipitate and beaked, 25-30 X 40-55 mm., with a sparse lacework of 

 fine fibers between the valves below in dehiscence: seeds black, very 

 dull, 6-7 X 8-9 mm.-Plates 18-28. 



Mexican table-land, from Tehuacan southward at least 

 as far as the southern part of the Tomellin Canon. 



Specimens examined:— ^pontoneows; Tehuacan, Trelease, 

 August 1903, about the northern foot of the Cerro Colorado 

 where it is the prevalent species of this group, called "espa- 

 diUa;" Rose & Rose, September 1906, no. 11263. Mexia, 

 in the Tomellin Canon, Trelease, February im.-€ulti- 

 vated: Missouri Botanical Garden, 1888, 1892, 1906. Munich 

 Garden, as A. pugioniformis, June 12, 1853. Palermo Garden, 

 Borzi, 1905,— an aberrant leaf with the open attenuate spme 



As th'^'^nly segregates that appear possible with present 

 material, may be noted:— 



Leaves of the nonnal form and size, entire. var. 



Leaves large and broad, repand. var. lan/oiui. 



The first of these is Mr. Baker's A. integrifolm, which pos- 

 sibly belongs to an entirely different group, and of which 

 nothing further seems now ascertamable unless the type, 

 which was still living in 1897, may exist at Kew. The second, 

 which perhaps is A. Besserermrut rmpr, and also A svdhury- 

 ensis of the Peacock garden that was made the type of A 

 carudnrm Baker, is known to me only from two large sterile 

 plants observed in February 1905 in the mountains behind 

 the Riego hotel at Tehuacan, and from a leaf sent from 

 Tehuacan by Dr. R. Endlich m 1907 (no. 1929). Its large 

 somewhat flexuous spine and subrepand margin suggest a 

 possibility of hybridization with the associated A. Verscfmf- 

 feltvi A plant cultivated at Fairmount Park as A. macra- 

 cantha longifolm, though clearly of this species, is peculiar m 



