NOTES ON SPECIES OF LOTUS § PEDBOSIA. 383 



acute, broader below, distinctly subpetiolate. Peduncles very short, 

 hardly exceeding J in. in length, 1- to 3- or 4-flowered (in tlie 

 specimens before me nearly all 2-flowered). Flowers yellow. 

 Bract normally trifoliate. Calyx clothed with adpressed hairs, 

 subbilabiate, teeth subulate, about as long as the tube. Style with 

 a short but very distinct tooth. Stigma capitate. Pod about f in. 

 long, straight, rather thick, more or less strangulate from abortion 

 of seeds, which are about twelve to fifteen in number in the speci- 

 mens before me. 



Salvages (Grand Piton) ; apparently common. Collected by 

 Ogilvey Grant, Esq., iu May, 1895. 



Very distinct from Pedrosia Paha, judging from Lowe's de- 

 scription and from the specimens (which are in very poor condition) 

 at Kew. I should have referred these to ordinary L. glaucm. No 

 doubt L. salvagemis comes also very near that species, if taken in 

 an extended sense, but even so it seems to be sufficiently distinct 

 to be kept apart. In the structure of the stipules it approaches 

 Pedrosin nerjlecta Lowe, but diflers markedly iu the extremely short 

 peduncles, a character in which it resembles P. Paivte, and which 

 led Lowe (in the absence of flowers) to surmise that the last- 

 mentioned form belonged to the group of which L. macranthus 

 Lowe may be taken as the type. But I can see no valid ground 

 for this suggestion. 



Lotus sessilifolius DC. DeCandoUe remarks that this species 

 is allied to L. fjlaucus, but differs by its perennial habit, shrubby 

 stem, longer and narrower leaves, larger and more numerous 

 flowers, and cylindrical pods. I doubt whether any of these 

 characters can be absolutely depended on except the perennial 

 habit and somewhat shrubby stems. The leaves of L. ylaucus var. 

 an<jmtifolius sometimes hardly differ from those of L. sessilifolius 

 except in being stalked, while I have occasionally, though very 

 rarely, found plants of L. sessilifolius with a few of the leaves with 

 a distinct though short petiole. The floral characters are still less 

 to be depended on. However, the habit and general appearance of 

 the plant is very different from that of L. glaucus, and I have no 

 doubt that it is a good species. It is quite confined to the Canarian 

 archipelago, occurring, always in maritime or submaritime situa- 

 tions, in Tenerife (Guimar, S. Cruz, S. Juan de la Rambla). It is 

 also reported from Hierro (Bourgeau, 798) and from Gomera (near 

 S. Sebastiano), but these plants require further study. The plant 

 from S. Juan de la Rambla is remarkable for the very silky-villous 

 clothing of the calyx, and for a tendency to produce a certain 

 number of more or less spathulate leaves. 



Lotus lanzerottensis Webb & Berth. Another species of 

 the fjlaucus group, but readily distinguishable from L. (jlaucus by 

 the broad obcordate leaflets and roundish or obcordate-rhomboid 

 subpetiolate stipules, recalling those of L. salvac/ensis. The clothing 

 of the leaves recalls L. sessilifolius. 



Canaries : only in the islands of Lanzerote and Fuerteventura, 

 from the latter of which a var. villosa is recorded in Phyt. Can. 



