218 leaves, sessile or shortly nedicillate, bracteate. Calyx- 5-6- 

 lobed, sonetimes produced above its union vath the ovary into a 

 short tube or cup. Petals numerous, free, linear. Stamens numer- 

 our, erect, and v/hen the calyx has a tube above the ovary, arising 

 from the ton of that bue with the petals; sterainodes none, ^tig- 

 mas 9-15, filiform; no style. Ovary inferior, 9-15-celled; pla- 

 centas dovn the centre of the outer wall of the cells and often 

 extending tiart of the way ut) the central axis. Capsule (I'ig. 104) 

 subhemispheric, with 9-15 valves and cells; each valve with a pair 

 of separated, parallel, expanding keels often ending in awn-like or 

 membranous points and with membranous wings at the sides; cells 

 acutely roofed v.dth rather rigid dell-v.dngs or separated elements 

 of the cell-partitions with their outer ends turned back so as to 

 form a somewhat trximnet-shaped onening to the cell, which is partly 

 or nearly closed by a small or Isrge tubercle vcell-wings flat and 

 membranous and the tubercle rudimentary in ^, Roodiee); the ridges 

 formed by the roofs of the cells together with their turned-back 

 ends form, a series of small cavities alternating with the cells, 

 the whole being raised in a sort of corwn above the level of the 

 expending-keels. Seeds somevrhat compressed-ovoid and a slightly 

 anguler from mutual pressure, with a nipple at one end. 



Species 5, natives of the ^^roo region of South Africa, the 

 type of the genus being P. magninunctata K, S, -t^r. 



The name is derived from the Latin, punctilltom, a dot, in 

 allusion to the conspicuously dotted leaves. 



This genus is closely related to Pleiospilos and, except in 

 vegetative characters, is tecbinically only to be distinguished from 

 that genus by the opening to the cells of the capsule being nearly 

 closed by a large tubercle. But this tubercle is doubtless of some 

 functional importance to the plants havng it, and as it is combined 

 vrith a different type of foliage, ^ consider it to be cuite as im- 

 portant a generic character as, for instance, the presence or 

 absence of scales upon the claws of i::;etpls or ^t the mouth of the 

 tube of a corolla are held to be. The function of these tubercles 

 is verj'' obscure; they occur in several distinct genera, as so far 

 as I have been able to discover, they prevent, rather than aid, the 

 dispersal of the seeds. "^ej offer a problem that requires inves- 

 tigation, for ^^ature would scarcely develop such pronounced struc- 

 tures as these tubercles are unless they were to be of some impor- 

 tance to the plant. As ^ have previously states, the fruits of 

 these plants are the most complicated in the world of flowering 

 plants and the most puzzling to understand, 



KEY TO IHE SPECIES. 



1,— Leaves less (or not more) than tv.'ice as long as broad, 

 l-2i inches long, 8-14 lines broad, flat on the face, 

 very obtuse, flowers two inches in diameter, sessile. 



5, Roodiae 



Leaves three to four (or more) times as long as broad. 2 



2, — Leaves 6i-ll lines broad or perhaps sometimes broader, if- 

 3 inches long, flat or faintly concave above, very 

 conspicuously marked with large dots on a glaucous- 

 green or brownish-tinted ground; flowers sessile, 2-2-| 

 inches in diameter. 1. magnipunctata 



Leaves 3^-7 lines broad, if very conspicuously dotted 



