40 ON THE REDISCO^-ERY OF THE GENUS EUSTEPHIA. 



Aud again, at page 156, he returns to the charge at greater 

 length. The foundation of his positiveness is probably the fact 

 that the plant bears a considerable resemblance, in flower and 

 general habit, to the Chilian Phycella ignea, in which there are 

 minute teeth at the base of each filament. However that may be, 

 it turns out that the " mendosissime' existed in his imagination 

 only, for during the last few years it has reappeared in English 

 gardens, and the substantial accuracy of the plate and generic and 

 specific description of Cavanilles is fully vindicated ; and I now, 

 therefore, claim again for it an admitted rank as a distinct 

 monotypic genus. 



What I know about its reappearance is as follows : — I received 

 in May, 1875, a specimen from Mr. Green, of Eeigate, who was 

 formerly gardener with Mr. Borrer, and afterwards with Mr. Wilson 

 Saunders. All that he was able to tell me of its history was, that 

 he had bought in London in a miscellaneous lot of Andine bulbs. 

 Not thinking of looking in the dim limbo -land to which Enstejjhia 

 had been consigned, I failed to identify the plant, and published it 

 in the ' Gardeners' Chronicle ' as a new section of Phcedranassa, 

 under the subgeneric name of Odontopus. Since that time we have 

 received a li\ing plant at Kew from Colonel Trevor Clarke ; and on 

 looking lately accidentally at the figure of Cavanilles, whilst 

 referring to something else, I found that my Phcedranassa rubro- 

 viridis was identical with the long lost Eustephia. And on looking 

 amongst the unnamed Amaryllidacew in the Kew herbarium, I 

 find that we have two specimens gathered by Maclean in Peru ; 

 and another specimen, sent by Herbert from his garden at Spof- 

 forth, shows that this is the same plant which he mentioned, by 

 name only, under tab. 3865 of the 'Botanical Magazine' as 

 Eustephia Macleanica. So that now we have its whole history 

 satisfactorily cleared up. 



There is another genus of AmanjllidacecE — Calliphruria, of which 

 the type, C. Hartiregiana, has lately been introduced into cultivation 

 by Mr. Bull and figured in the * Botanical Magazine, (tab. 6259) — 

 that has petaloid tricuspidate filaments. Eunjcles also often, and 

 even Eucharis sometimes, has the corona slit down to the base into 

 six distinct divisions. The five genera abeady mentioned, Phycella^ 

 Eustephia, Calliphruria, PJurycles and Eucharis, furnish, in the 

 order in which I have just mentioned them, a gradual transition 

 in stiiicture between the typical Amaryllidacea., like Amaryllis and 

 Crinum, with free filiform filaments, and the monadelphous 

 PancratiecD, like Pancratium, Ismene, and Hyvienocallis, in which 

 the filaments are united into a com^dete corona in their lower part. 



The following description is drawn up entirely from the 

 specimens I have myself examined, and may be compared with 

 the diagnosis and plate of Cavanilles as a test of the accuracy of 

 my identification. The only jDoints in his description to which I 

 should be at all disposed to take exception are, that he speaks of 

 the bracts as a calyx, of the j)eriantli as a corolla, describes the 

 part of the said corolla (in the generic, but not in the specific 

 description) as five instead of six, and figures six distmct hollows 



