Johnson. — -On A rceuthobium Oxycedri . 143 
cavity present, and the apical central sac shown nearly en- 
closed in cellular tissue is not the embryo-sac but the embryo 
imbedded in endosperm, as is the case in fig. 12 of the same 
Plate. 
2. The Ovarian Papilla in other Loranthaceae. 
In 1836 Griffith’s 1 valuable and interesting paper, ‘Notes on 
the Development of the Ovule of Loranthus and Viscum ’ ap- 
peared. In Viscum [species not given] he found a nipple-shaped 
process in the ovarian cavity, at the base of which he saw two 
projecting more or less pendulous bodies, which he regarded 
as naked ovules, the nipple- shaped process being the placenta. 
This arrangement was so different from anything seen in any 
genus of the Loranthaceae, more especially so different from 
what several subsequent observers saw in species of Viscum , 
and at the same time so like the condition of the gynaeceum 
in the Santalaceae (e. g. Sant alum album as described by 
Griffith 2 himself), that by Hofmeister and others Griffith’s 
Viscum was regarded as a genus of the Santalaceae. Against 
this, however, Treub has protested, for he considers our 
knowledge of the structure of the gynaeceum of the Loran- 
thaceae too imperfect to permit of dogmatism, and his protest 
is supported by the structure of A. Oxycedri. It is not 
difficult to see how the two apical embryo-sacs buried in the 
nipple-shaped process in A. Oxycedri could be derived from 
the two basal pendulous projecting bodies on the nipple- 
shaped process in Griffith’s Viscum . In Griffith’s Loranthus 
Scurnda neither ovarian cavity nor papilla was seen. In 
a later paper — read 1843 — Griffith 3 describes a Malaccan 
Viscum in which he found no nipple-shaped process, but 
such a process was found in Loranthus bicolor , though in it 
subsequently rendered obscure by the migration of the embryo- 
sacs half way up the style. 
1 Griffith, in Trans. Linn. Soc., xviii. p. 76. 
a Griffith, in Trans. Linn. Soc., xviii. p. 59. 
3 Griffith, in Trans. Linn. Soc., xix. 
