of the Family o/"Burmanniacese. 379 



the dehiscence is effected by the rending of the thin membrane, both of the 

 disc and the walls of the cell, in the manner shown in Tab, XXXVII. fig. 1,/. 

 and fig. 2, /. of the eighteenth volume of the Society's ' Transactions.' 



If, again, we imagine the stamens of Apteria with their curiously-expanded 

 connective, retroflexed at their origin, we should see a structure of these or- 

 gans quite analogous to that existing in Ophiomeris. Such considerations, 

 combined with other manifest characters, serve at once to clear the way for 

 the admittance of these plants into Burmanniacece, in which natural order I 

 propose to class Ophiomeris and Thismia, under a distinct section, to which 

 the name of Thismiece may be given. This family would hence be subdivided 

 as follows : — 



BURMANNIACE^. 



1. BuRMANNiE^. Perianth, triptenim. Stam. 3. Ovar. 3-locul. Placent. centrales. 



Capsula longit. dehiscens 1. Burmannia. 



Capsula transv. fenestrata 2. Gonyanthes. 



2. APTBBIB.S. Perianth, exalatum. Stam. 3. Ovar. 1-locul. Placent. 3 parietales. 



Capsula irreg. 3-valvis 3. Dictyostega. 



Capsula lateral, hians 4. Cymbocarpa. 



Capsula apice 3-valvis. Stamina appendiculata . . .5. Apteria. 

 Capsula irreg. dehiscens. Petala nulla 6. Gymnosiphon. 



3. Thismiece. Perianth, exalatum. Stam. 6. Ovar. 1-locul. Placent. 3 parietales. 



Pericarpium circumscissum. 



Tub. regul. Stam. monadelph 7. Thismia. 



Tub, gibbus. Stam. omnin6 libera 8. Ophiomeris. 



Having before alluded to the fact suggested by Mr. Griffith of the analogy 

 which these two genera bear to Tacca, I will now, before making any obser- 

 vation on this head, first point out the remarkable resemblance in habit and 



into the cavity of the cell ; but whether the above-mentioned process formed the communicating me- 

 dium of impregnation from the stigmata to the ovules, or whether it was foreign to its structure, I had 

 no further means of ascertaining. I feel disposed however to prefer the former view, as it would seem 

 to bear some analogy to the 2 glandular lobes that I formerly observed beneath the disc upon the ter- 

 mination of each of the 3 placentary ribs in the capsules of Dictyostega, Cymbocarpa, and Apteria, 

 and which are represented in Tab. XXXVIII. fig. 4, m, illustrative of my memoir above-cited. If 

 this should be the case, it would become a matter of some interest to ascertain what function these 

 glandular masses perform in the process of fecundation, for it is natural to conclude, from their con- 

 stant presence, that some definite office is assignable to them. 



