and Affinities of Tr a fella. 115 
Fig. 63 b. Similar outline of leaf, intermediate in form between a submerged and 
floating leaf. Nat. size. 
Fig. 63 c. Two teeth from a submerged leaf seen under a low magnifying 
power, so as to show the course of the vascular bundles as they run to the water- 
glands, w. g. 
Fig. 64. Epidermis of submerged leaf, showing four-celled glands, ep ., epi- 
. dermal cells; /., pedicel of gland (dotted outline). 
Fig. 65. Vertical section of same gland, i. s., intercellular-space ; p ., pedicel of 
gland. 
Fig. 66 a and b. Pretrea ; four-celled gland from leaf, seen from above, and in 
section, for comparison with Figs. 64 and 65. 
APPENDIX. 
(1) Since this paper has been in proof, Dr. Vines has 
kindly called my attention to the similarity presented by the 
sucker or appendage described above and the cotyloid cell of 
Avicennia ( vide M. Treub, in Annales du Jardin Botanique de 
Buitenzorg, vol. iii, 1883, p. 79). The cotyloid cell resembles 
the ‘appendage’ of Trapella in being a ‘feeder,’ and also in 
its origin — in that it originates outside the embryo-sac. 
Whether the cotyloid cell be a cap-cell or not remains to be 
shown. The cotyloid cell is seen, after the commencement 
of the development of endosperm, lying beside the embryo-sac 
(loc. cit. PI. XIV, Fig. 1 8), to which it may be a sister cell. 
This remains to be shown. 
(2) In one of his letters, Dr. Henry mentions that in the 
Chinese illustrated Botany ‘ Chi Wu Ming,’ vol. xvii, folio 43, 
an aquatic plant somewhat resembling Trapella is figured. 
This plant is known as ‘Chi-Mi’ in the province of Chi-li, 
and ‘ Ch’a ling’ (i.e. Tea Trapa) in the province of Honan. 
The description is too incomplete for any decision as to its 
affinity to be made, and the figure (of which Dr. Henry 
encloses a tracing) shows a plant with habit somewhat re- 
calling Trapella , but with leaves alternate, and pseudopodial 
branching. 
