166 A NEW KEY TO THE GENERA OF AMARYLLIDAGE^. 



REMARKS. 



1. Galanthi(,s. — Here I consider there are three species, viz., 

 1, nivalis, of which I)nperati,'Bevt., iatif alius, Enprecht, and reflexus, 

 Herb., are varieties ; 2, FAtvesii, Hook, fih, in ' Bot. Mag.,' t. 6166 ; 

 and 3, plicatus. G. PuyimF-O/gfs, of Orphanides, I know by name 

 only. 



2. Leucqjum. — Inchiding Erinosma and Acis, as in Kimth, good 

 as subgenera, and under the latter Buwinia, of Parlatore, — species 

 8 or 9. L. Hernandezianum , Camb., is evidently the same as 

 L. jiuJchellum, Salisb., which is common in London gardens. Of 

 the species of Acis in Kunth, I should join grandifloriim with 

 trichophijlhim. None of those figured by Jordan and Fourreau 

 (' Icones,' figs. 103 to 108) seem distinct from those given in Kunth. 

 L. autumnale is autumnal, all the others vernal. L. hyeniale is a 

 misleading name, as the plant flow^ers in April (not autumn, as 

 stated by Kunth); Paiminia niceceensis, Jord. & Fourr., is identical 

 with it. 



3. Strumaria. — Includmg Imhojia. Seven sj)ecies, Baueriana 

 being conspecific with linguafolia, and BurclieUiana and Bergiana 

 with geminata. 



4. Haglockia. — Monotypic. 



5. AjmdoUrion. — Three species, as described 'Journ. Bot.,' 

 1878, p. 74. 



6. GethgUis. — Four species known in flower, and one clearly 

 distinct, and one doubtful m fruit only. G. acaulis, Blanco, is 

 doubtless a CurcuUgo. 



7. Cuoperia. — Two species ; mexicana, now known in flower, 

 being identical with Drummondii. 



8. Sternhergia. — Including Oporanthus ; five species, Kunth's 

 needing much reduction. S. dalmatica, (Ptnmsis, citrina, and 

 perhaps even Clusiana seem to be the same species as colchicifiora, 

 and Fischeriana and probably e.i-igua to belong to lutea. Of new 

 discoveries we have .S'. macrantha, J. Gay {S. latifolia, Boiss.), the 

 finest plant in the genus, and 5'. pidcliella, Boiss. & Blanche. 

 S. sicula, Tineo, is a variety of lutea, and S. e.cscajja, Tineo, alto- 

 gether doubtful. 



9. Zephgranthes. — Two subgenera, as in Kunth. Species about 

 fifteen, Grahamiana being a form of sessilis ; Jfavesceiis and acuminata 

 of viesochloa; Lindlegana and nervosa, M. & G., identical with 

 pcdUda ; and texana the same as Habranthus Andersoni var. 

 texanus. 



10. Fyrolirion. — The three species in Kunth are doubtfully 

 distinct ; albicans rests still solely on the authority of Feuille ; and 

 there are two novelties still undescribed in Mandon's plants of 

 the Bolivian Andes. 



11. Haimanthus, — Species about thirty, five of which are 

 Tropical African, and the rest Cape. The structm-e of the flower 

 is very uniform through the genus. Of the plants in Kunth, 

 concolor and Hookerianus are probably the same as sanguineus. 



