Mr. J. Miers on the Affinities of the Olacacese. 179 



in the female flowers, also 3- or 4-lobed, the ovarium is half im- 

 mersed in a similar adnate fleshy cupshaped disk ; it is 1 -celled, 

 with three ovules suspended from a free central placenta; the berry 

 contains a single naked seed, enclosing a compressed heart-shaped 

 albumen, with a minute embryo in its almost cordate summit ; 

 the radicle is terete, the upper moiety of which is nearly exserted, 

 having only a thin pellicular albuminous covering ; while its lower 

 moiety, and two exceedingly diminutive cotyledons, are imbedded 

 within the substance of the albumen, in the marginal sinus. 

 These characters are so perfectly distinct from the Loranthacece, 

 that it appears to me the genera above mentioned should form 

 either a separate family (the Viscacece), or be considered as a sub- 

 order of the Santalacece. The only points of resemblance between 

 Viscum and the Loranthacece are, the position of the stamens 

 opposite the lobes of the corolla or perianthium, the manner of 

 development of their seeds, their glutinous properties, and their 

 parasiticism, characters equally possessed by other families : they 

 are certainly quite distinct in habit. Mr. Griffith states, that 

 the Indian species of Viscum have three ovules suspended from a 

 central column, thus agreeing with the Brazilian species, which 

 I have called Allobium. The ovules of Viscum album are said by 

 M. Decaisne to be erect, but I have elsewhere offered reasons 

 why we may infer that they are in reality suspended, and only 

 apparently erect, as in Champereia, &c. 



In first pointing out the affinity of the Loranthacece with the 

 Santalacece, many years ago (Prodr. 352), Mr. Brown probably 

 had Myzodendron and Viscum in view, as at a later period (Linn. 

 Trans, xix. 232) he has alluded more distinctly to the similarity 

 in the construction of the ovarium of the former genus with that 

 which forms a pecular feature in the Santalacece. In indicating, 

 on the other hand, the relation of the Loranthacece witlf the Pro- 

 teacece (Flind. Voy. App. 549), the same distinguished botanist 

 probably had only Loranthus in consideration. The evident 

 affinity of Viscum, just mentioned, was also remarked by Prof. 

 Decaisne, in his memoir on the pollen of that genus, before 

 cited, on comparing the ovules of Viscum album with those of 

 Thesium. Brongniart (1843), adopting this view, arranged the 

 Loranthacece in a separate class, with the Santalacece and Ola- 

 cacece. The same affinity between these three families (at least 

 as far as regards Myzodendron and Viscum) has since been con- 

 firmed by Dr. Hooker, in his very able investigation into the 

 relations of the former genus (Flor. Antarct. 293) ; and the 

 strongest evidence in proof of this affinity is given in the com- 

 plete analysis of its ovarium, from its early development to the 

 perfection of the fruit, the details of which are there exemplified 

 in plate 104. fig. 10 to 20, and plate 105. fig. 12 to 21. 



12* 



