Mr. J. Miers 07i the Affinities of the Olacaceae. 177 



to diflPer little from the Aquifoliacece. The word torus is gene- 

 rally confined to that fleshy termination of the peduncle in the 

 bottom of the calj^x seen in Ranunculacea, and more especially 

 developed in such orders as the Anoiiacea, Magnoliacece, &c., but 

 when it rises in more varied or determinate shapes, it takes the 

 name of hypogynous glands, annular ring, flat, pulvinate or 

 cupuliform disk, &c., according to the pecuHar form it may 

 assume, or the position in which it is engendered. 



The epigjnous gland, so highly developed in Schiipfia, Arjoona, 

 Cathedra, and other genera of the Olacacea, is an equally con- 

 stant feature of the Santalaceee, where in Ejcocarpus aphyllus it 

 is largely and prominently seen in the form of a 4-lobed cushion, 

 broader than the summit of the ovarium, which is almost entirely 

 superior ; this is quite independent of its hypogjnous disk, which 

 is also present as usual in the family. This organ, whose exist- 

 ence I first pointed out in Hyoscyamus, I have since found to 

 occur frequently upon the summit of a superior ovarium. 



This inquiry into the affinities of the Olacacea has led to an- 

 other conclusion of some interest. In my memoir upon Cathedra 

 {huj. op. vol. vii. p. 45J:), while describing its cm'ious anthers, I 

 pointed out a very analogous structure in Choretrum and other 

 genera, mentioning at the same time a similar formation of the 

 anthers in Myzodendron, so beautifully illustrated in the ' Flora 

 Antarctica ' by Dr. Hooker, who has there also given the analysis 

 of its ovarium and fruit, proring by indisputable evidence its 

 relation to the Santalacece and Olacacea. I will now endeavour 

 to show, that neither this genus, nor Viscum, bear any relation 

 to the Lorardhacea, where they have been placed by almost every 

 botanist. Tlie genus Viscum has been a fi-equent subject of in- 

 vestigation by many eminent physiological botanists, and Richard 

 first described the very remarkable structure of the anthers of 

 Viscum album, of which we find no parallel formation : these are 

 well represented (Ann. Mus. torn, xii. tab. 27) as being com- 

 posed of ver)' numerous cells, each containing distinct aggrega- 

 tions of polleu-gi'aius, and which burst open and discharge their 

 contents by the rupture and contraction of the vesicular tissue 

 that covers their surface ; in this respect it bears no resemblance 

 to the structure of the anthers of Myzodendron. On the other 

 hand, upon examining the anthers of the Brazilian species of 

 Viscum, I find their structure quite opposed to that described in 

 V. album, and somewhat analogous to those of Myzodendron ; 

 they ai'e 2-lobed and subcordate, approaching much the form of 

 those of Cathedra ; they are quite distinct and free from the lobes 

 of the perianthium, are nearly sessile, and consist of two parallel 

 cells, enclosed in thick crystalline walls, as described in that ge- 

 nus, and appear to discharge their fertilizing power in the same 

 ambiguous manner by two covered pores in the apex : the polleu 



Ann. i^ Mag. N. Hist. %tv.^. Vol.\n\. 12 



