curved not ftraight, approaching an oval not a cuneate form, 

 points though obtufe not fo broadly rounded ; colour light 

 yellow. Strong plants from old bulbs generally have two 

 branches ; flowers, three to five, fcentlefs, or, according to 

 Mr. Curtis, (lightly fragrant; leaves, as in its relatives, 

 more numerous than is general in Ixia, fix to ten, outer ones 

 gradually Ihorter, outermoft fcarcelv half an inch long. De 

 la Roche's variety is blotted and {beaked with brown ; but 

 if Ixia /acem, which we have not yet met with, be a diiiinft 

 fpecies, this would rather be a variety of that. 



Propagates fall by feeds, ftem-bulbs, and root-bulbs, and is 

 become one of the commoner! fpecies in our gardens, yet has 

 never before been figured, except in the Botaniit's Repofitory. 

 Preferves its colour, when dried, unimpaired. Cultivated by 

 Miller in 1758. 



It has been fuppofed that many remarkable varieties in this 

 genus have been produced by the induftry of Dutch gardeners, 

 but this is very doubtful, for certain it is that fcarcely a va- 

 riety, or what is fuppofed to be fuch, has been imported from 

 Holland, that has not been produced alio from bulbs that 

 were undoubtedly gathered wild or difcovered in fpecimens 

 collected at the Cape. Whenever therefore a plant affords 

 marks of distinction from its congeners, exceeding thofe al- 

 loted to variety by Linn£us, and is not known to have been 

 raifed from the feeds of the identical fpecies of which it is 

 pretended to be a variety, we, without hefitation, record it a 

 diftinct fpecies : as ftudioufly avoiding, on the other hand, to 

 enumerate as fuch, thofe that merely differ within the rules 

 prefcribed by our great mailer. When, as in this inftance, 

 we are but flightly acquainted with the local economy and 

 precife habitat, this rule requires to be ftrictly attended to. 



N. B. In the enumeration of the fpecies of Watsonia (fee 537) 

 we omitted Ixia fpicata (fiflukfa, Bot. Mag.) and Ixia plantag'nea, 

 which we now propofe for infertion. — Watsonia plantaginea will W 

 ihortly given. 



