Mr. J. Miers on the Calyceracese. 177 



seated upon the free portion of the filamentous ring, above the 

 point of its insertion upon the corolla. 



In regard to the prominent tubercle which crowns the ovary 

 in all the CalyceracecBj which Richard held to be a disk, all that 

 I have seen tends to a conclusion at variance with the precept of 

 Mr. Brown, which attributes to the flower in this family, as its 

 essential character, " the absence of an epigynous disk," and, as 

 " a necessary consequence " of this character, " the accretion of 

 the base of the style to the base of the corolla." We have strong 

 presumptive evidence of the actual presence of an epigynous 

 disk, analogous to that in Composite ; and it may be inferred 

 that the u accretion" above mentioned arises from its intervention 

 and confluence, both with the base of the style which it sur- 

 rounds, and with the tube of the corolla, over the inner surface 

 of which it is spread to a considerable extent, so that when the 

 corolla falls off we always find the bottom of the tube closed, as 

 if by a plug, owing to its presence. The nature of the five 

 areolar glands is somewhat connected with this consideration : 

 they are always alternate with the stamens ; and their origin may 

 be attributed to two sources : either they may be viewed, as 

 Mr. Brown regarded them, as a row of additional abortive sta- 

 mens ; or they may be considered as glandular protuberances 

 appertaining to a disk extended, in the manner above mentioned, 

 over the inner surface of the tube of the corolla. The former 

 view will not stand the test of analogy, when compared with the 

 numerous and varied developments seen in the neighbouring 

 order Composite; and the latter suggestion may be adopted 

 without having recourse to the idea that the glands are "a series 

 of modified stamens." In regard to the existence of such a disk 

 as I have suggested, we have some evidence in Nastanthus, where 

 we find the usual confluence of the lower half of the staminal 

 filaments into a monadelphous tube {tubillus of Richard), which 

 soon becomes adnate to the tube of the corolla : we find in the 

 mouth of this " tubillus " an internal row of minute, free, linear 

 hyaline teeth, alternating with the upper or free portions of the 

 filaments, rounded at their apex, and connected together by an 

 acute sinus ; they evidently form an inner whorl with the fila- 

 ments, and do not intervene between them. These teeth bear 

 more the semblance of abortive stamens than the areolar glands; 

 but the same reasoning which leads to the rejection of such a 

 nature for the one will deny it to the other. We may with 

 greater probability consider these teeth as constituting the 

 margin of such an adnate disk as I have suggested, the presence 

 of which is supported by much collateral evidence ; and we may 

 moreover, with great reason, attribute to the existence of this 

 disk the cause of the confluence of a portion of the filaments 



