13 alternate v;ith the valves, each valve bears upon it (in the v;ing- 

 like membranes on its inner face) the partitions belonging to the 

 two adjacent cells. The cells themselves are very narrow, end 

 each contains about one to four seeds. 



All the species belonging to this genus are glabrous, except 

 on the petals and stamens, and all grov.' i pure and or in sandy pla- 

 ces. Therefore they should be cultivated in a light sandy soil, 

 and the soil not allowed to get too dry at the lower part of the 

 pt. They are best watered from below by standing the pot in a 

 saucer of water. Their f louvers are very shov.'y. But I have seen 

 one species growing and flov>rering most luxuriantly in ordinary 

 garden soil in the onen air in south Devonshire. And in climates 

 not too cold for these plants to be cultivated in the O' en air, 

 the species of this genus ought to find favour among cultivators. 



The herbarium material that I have seen is very imperfect, 

 the root, main stem and central tuft of leaves usually being 

 euite absent, while the flowers are mostly badly pressed, and, 

 ripe fruit definitely belonging to the same plant as the flowers 

 is usually quite absent. So that dreid material is difficult to 

 determine, and as the published figures of the species of thus 

 genus have been indiscriminately named, much confusion of nomencla- 

 ture hpve been made, as no heed has been paid to the fev/ charac- 

 ters which the descriptions and figures portray. These characters 

 I have made use of to indicate the various species in the key that 

 follows. One of the be-^t characters for distinguishing the species 

 is that of the fruit, but this character I have not been able' to 

 made use of, as it is known to me only in a fev cases, for the 

 collection of the fruit in definite connection with flowering spec- 

 imens appears to have been entirely neglected. 



From the study I have made of this genus, I belie?:e that it 

 v;ill ultimately be found to contain several other species as yet 

 undescribed. I have fruits only of two others that are certainlj'- 

 distinct from any here described, and there are probably others, for 

 in general appearance the species are so similar that their very 

 real distinctions h^ve not been noticed. For example, I find 

 that plants with fibrous roots and a distinct stem, under the central 

 tuft of lepves, and others with a tuberous root-stock and a sessile 

 central tuft of leaves have indiscriminately been named in South 

 Africa as beins "M. pugioniforme Linn.," which at the present day is 

 unknown to botanists. 



In the synonymy of the species, references pre to previous 

 monographs are sometimes omitted, because I find that as therii 

 ouoted ther is 

 14 sometimes a misture of snecies that mskes the reference worthless 

 and so it is not quoted. 



Key to the Species. 



1. Roots fibrous; plants developing a single and distinct 



stem 2-3 inches to 2 feet or more high under the central 

 tuft of leaves, erect or bent do^v-n from the weight of 

 the leaf-tuft end flovering branches; leaves sharply 

 triangular in cross-section. 



^ootstock tuberous, fleshy, long and cylindric or spindle 

 shaped or irregularly ovoid, producing either a tuft of 

 leaves without a stem under it, froia among v/hich the 



