28 



retained for the garden one seed only has germinated, so that 

 the species is probably of extreme rarity. It is a handsomer 

 looking plant than GarTya elliptica, with oval laurel-like 

 leaves, which are covered with down on the under side. 

 According to Mr. Hartweg it grows on the mountains in the 

 northern provinces of Mexico. He found it near Guanaxuato 

 a shrub from fifteen to eighteen feet high ; but at Anganguco 

 it formed a tree with a trunk two feet in diameter. 



54. ARISJilMA macrospatha. Bentham plant. Hartweg. p. 52. no. 394. 



A small stemless plant, with tuberous roots like those of 

 the common Arum, purple-stalked pedate leaves, and a pink 

 or purple spathe from five to six inches long. It will pro- 

 bably form a greenhouse herbaceous plant. Mr. Hartweg 

 found it in shady woods near Morelia flowering in July, and 

 sent it to the Horticultural Society, with whom it has been 

 raised. 



55. HYMENOCALLfS rotata. Herbert Amaryll. p. 217. 



There is a good figure of this plant in the Floral Cabinet, 

 Vol. 2. p. 51, and an interesting account of its peculiar habits 

 at p. 47 of the same volume, from which it appears that the 

 species lives in the deep muddy swamps in the neighbourhood 

 of Mobile, which it studs with its beautiful snow-white starry 

 blossoms, and perfumes with its fragrance, at a time when, 

 from the softness of the mud such places are unvisited except 

 by the most adventurous travellers. It is stated that the 

 bulbs frequently lie imbedded two feet in the mud, so that 

 the entire plant is full four feet high, and the collector, in 

 his efibrts to obtain the bulbs was frequently sunk above 

 his middle in this unpleasant bath. In the summer months, 

 when such swamps become hard, the plant dies down and dis- 

 appears, its very site being covered with other vegetation. 



It is much to be regretted that in publishing- an account 

 of this fine plant the authors of the work in which it ap- 

 peared should have stated it to be new, and named it Ismene 

 Knightii ; a double error being thus committed, the plant not 

 belonging to the genus Ismene, but to HymenocoMis, and not 

 being a new species, but one long since published in Messrs. 

 Loddiges' Botanical Cabinet, t. 19, and in the Botanical 

 Magazine, t. 827, under the name oi Pancratium rotalum^ by 

 which it is well known in collections. These are the new 



