456 Mr. J. Miers on the genus Cathedra. 



other families, although they are generally considered to have 

 bilocular anthers ; in these instances the valves are much thinner 

 and reticulated in texture ; they are the same in Psittacanthus ; 

 but in Struthantlms, though the walls are crystalline, and as thick 

 as in Cathedra, they yet open in the manner just described. 



The comparison of the characters of Cathedra with other ge- 

 nera has led me into a general examination of the Olacacea, San- 

 talacece, and other allied families, in the course of which I have 

 met with numerous interesting and novel facts^ and from the 

 materials thus collected, I propose to give at an early period, a 

 review of each genus belonging to these orders in succession, 

 together with illustrated details of their characters. In the fol- 

 lowing memoir on Liriosma, I will offer some of the views I have 

 in consequence been led to adopt regarding the affinities of the 

 families above alluded to. 



The singular development of the very remarkable free cupuli- 

 form disk in Cathedra, that supports on its margin the petals 

 and stamens, is an important feature, as it serves clearly to de- 

 monstrate the true nature of the same organ, which, with few 

 exceptions, in all true Olacacece, is always more or less adnate 

 with the ovarium and quite free from the calyx, but which in 

 the genus Liriosma is connate with the calyx and wholly free 

 from the ovarium ; while in those genera of the Santalacece, 

 where the calyx and corolla are confluent in one common pei'i- 

 gonium, the disk is almost wholly coadnate with the latter : to 

 this feature therefore, as it is developed in Cathedra, frequent 

 recuri'ence Avill be made when we come to consider the different 

 genera in the manner I have proposed. The ovarium in the 

 plant under consideration is turbinate and flattened, the lower 

 moiety being smooth and concealed within the free surrounding 

 cup just mentioned, the upper moiety being covered with a thick 

 fleshy gland, in shape like a very depressed cone, marked with 

 twelve raised radiating striae; it is 2- celled at base, unilocular at 

 its summit, with a single ovule in each incomplete cell, suspended 

 from the apex of the free axile placenta. This structure is quite 

 analogous to that found in many genera of the Olacacece, and 

 different from what exists in Myrsinacece. I regret very much 

 that I did not meet with the ovarium further advanced towards 

 maturity, but this deficiency is in some degree supplied by 

 another very similar plant, evidently congeneric with the above, 

 and to which our attention was called nearly eight years ago by 

 Mr. Bentham (Lond. Journ. Bot. ii. 375), who says, " No. 5380 

 bis, of Gardner, from a single straggling shrub, found in a forest 

 at Tejuca, fourteen miles from Rio de Janeiro, is a very singular 

 plant, apparently allied to Olacacece, but unfortunately j)ast flower 

 in the specimens found. It has the habit, foliage, and inflores- 



