©ttstnal Mxtitlt^. 



A NEW KEY TO THE GENEEA OF AMARYLLIDACEM. 



By J. G. Bakek, F.L.S. 



In the present paper I propose to attempt to construct a key to 

 the genera of Amaryllidacea, in which flower-characters shall he 

 used as prominently as possible. The genera of the Order are 

 well and fully described in detail in three comparatively recent 

 works, viz., Herbert's ' Amaryllidacese,' published in 1837 ; the 

 fifth volume of Kunth's 'Enumeratio,' x^ublished in 1850; and 

 the fragment of Salisbury's ' Genera Plantarum,' which was issued 

 by Dr. John Edward Gray in 1866. But still, for everyday 

 working purposes, a key of the Order is greatly needed, partly 

 because it is very difficult to judge h'om a long description, where 

 nothing is emphasised, which characters are relied upon as 

 differential, and partly because the primary arrangement and 

 grouping used by these authors depends to a large extent upon 

 fruit and seed-characters, and these are seldom shown either by 

 living or dried specimens in the state in which they are commonly 

 submitted to botanists for identification. And another x^oint is, 

 that not to speak of Salisbiuy, whose ideas of a genus were such 

 that he proposes to make sixteen out of Xarcissiis as we commonly 

 understand it, many of Herbert and Kunth's genera are limited by 

 characters so faint that they are not at all likely to be adopted by 

 general botanists, who have the whole vegetable kingdom to deal 

 with. And one great good of these keys, whether they deal with 

 genera or species, is that they are so well adapted to bring into 

 prominence which of the separate individualities they deal with 

 rest upon an unsubstantial diagnostic foundation. As I have lately 

 dealt elsewhere with the AgavecB and Hypoxidacem, what I propose 

 to do now is simply to attempt to make a key for the Amari/llidacecB 

 proper and AlstrcemeriecB, and to follow it up by a few explanatory 

 remarks referring mainly to the reductions that seem needful in 

 genera and species as they stand in Kunth, and the comj)aratively 

 few novelties that have been discovered in the Order during the 

 last thirty years. 



Subordo I. Amakyllidace^ ver^. Herbae bulbosaB acaules, floribus 

 umbellatis vel solitariis. 



Tribus I. Galanthe^. Stamina epigyna,filamentisbrevibusliberis, 

 antheris ax3ice dehiscentibus. 

 I. Galanthus. Perianthii segmenta interiora exterioribus multo 

 breviora, cuneata, obtusa, profunde emarginata. Kicrojja, Asia 

 occidentalis. 



N. s. VOL. 7. [June, 1878.] y 



