Osyris, Loranthus, and Viscum. 193 



It is probable that the curious form of embryo-sac, so conspicuous in San- 

 talum and in Osyris, is of general occurrence in the natural family Santalacece. 

 I find, on referring to notes made in 1836, that a sac resembling that of Osyris 

 at its anterior, and that of Santalum at its posterior end, would appear to 

 exist in a Santalaceous genuspprobably allied to Sphcerocarya*. 



On the functions of the anomalous extension backwards I am by no means 

 inclined to advance any opinion. In establishing (or tending to do so) a 

 direct communication between the young seed and the vascular supplies, as 

 they diverge from the apex of the pedicel to their destinations, it may be 

 considered as analogous to the raphe of an ordinary ovulum. In its struc- 

 ture, however, and more especially the direction of its development, it appears 

 to be opposed to the usual form of raphe. It must also, I think, be remem- 

 bered, that in Loranthus, in which the same necessity of communication may 

 be assumed to exist, it does not appear to do so; and that instances are not, 

 perhaps, uncommon of ovula of ordinary form having no vascular or tubular 

 connexion with the vascular supplies. Neither must it be omitted, that in 

 Santalum, at least, there would appear to be a want of communication between 

 tfie ends of the extensions backwards and the vessels derived from the pedicel, 

 and that those of the barren ovula are equally, or nearly so, developed with 

 those of the fertile one. 



It would appear that these functions, whatever in reality they may be, do 

 not extend throughout the formation of the seed ; for not only does the pla- 

 centa seem to be torn up to a considerable extent from its original connexions, 

 but the tissue around it becomes altogether broken and dried up. 



From the consideration of Santalum and Viscum, I had been led to adopt 

 the opinion that the embryonary sac was probably the only essential part of 

 an ovulum. This opinion was formed on the apparent inutility of the nucleus 

 of Santalum, so far as its later functions were concerned, on the apparent 

 universality of existence of the embryo-sac, and on the structure of the Mergui 

 Fiscumf. And when I first examined the structure of Loranthus bicolor, I 



* I would recommend the prolongation of a membranous tube from the base of the ovulum back- 

 wards as one test of the affinities of Santalacece. 



f It is with reference to the determination of absolute essentiality that precise observations on the 

 origin of the sacs of Loranthus and Viscum are so necessary. Analogy would lead me to suppose that 

 some extent of nucleary base was indispensable. 



