60 JOURNAL, BOMBAY NATURAL HISTORY SOCIETY, Vol. VIII. 
A littoral species extending from the Mascarene Islands and Eastern Africa 
to Ceylon, the Andamans, Malaya, Australia and Polynesia; like Ochrosia 
borbonica this does not occur on the coast of India, though it is found as far 
north as Great Coco on the west and as Mergui on the east of the Andaman Sea. 
Meissner (DC. Prodr., xv, pt. 1, 262—264) omits to quote, and the Hora of 
British India (v, 188) does not cite Roxburgh’s account of Hernandia ovigera 
(Flor. Ind., iii, 577-578), which his own diagnosis clearly shows to be a species 
different from Hernandia ovigera Linn. (Aman, Ac., iv, 125), founded on 
Rumf’s figure (Herb, Amboin., iii, 198, t. 123) of Arbor ovigera. Roxburgh 
notes the discrepancies, and explains them by depreciating Rumf’s drawing. 
In reality, however, Roxburgh’s description is a most vivid and accurate one, 
made from living specimens, of the species named by Meissner (DC. Prodr., xv, 
pt. i, 263) Hernandia peltata. Roxburgh cites Gaertner’s figure (Pract. 1, 193, 
t. 40, f. 8) asa “ very accurate” delineation of the fruit of this tree—an exceed- 
ingly just remark, which, however, Meissner has overlooked, for he quotes 
Gaertner’s description and figure as referring to Linnaeus’ species, though they 
differ very materially from both Rumf’s figure and Meissner’s own description 
of the fruit of Hernandia ovigera, 
Hernandia peltata, the species now under review, is a purely old-world plant 
which has been treated by Linnaeus and, with the exceptions of Gaertner and 
Roxburgh, by all botanists subsequent to Linnaeus till the appearance of 
Meissner’s treatise (1864) as conspecific with the American Hernandia sonora ; 
even now Sir J. D. Hooker (Flor. Brit. Ind., v, 189) suspects that 4. peltata 
is no more than a variety of H. sonora. And the basis of the differentiation by 
both Gaertner and Roxburgh of the present plant from H. sonora does not lie in 
the differences between the two plants that Meissner has pointed out, but in the 
fact that Linnaeus included under H. sonora not merely the American tree to 
which Meissner would restrict that name, as well as the Ceylon tree which is 
undoubtedly H. peltata, but also—though doubifully and with the remark “ sed 
fructus alienus” (Amen. Ac., iv, 117)—the tree figured by Rumf (Herb. Amboin., 
ii, 257, t. $5) under the name Arbor regis. Believing, apparently, that Rumf’s 
Arbor regis was, a8 Linnaeus thought, a Hernandia—a belief perhaps partly just— 
but realising that it could scarcely be the tree he had before him, and seeing that 
it agreed so thoroughly with the figure and description of H. ovigera given by 
Gaertner, Roxburgh, not having in his possession specimens of the true #. ovigera, 
followed Gaertner in bestowing that name on this species. This course was hardly 
just to Rumf if Gaertner and Roxburgh believed Rumf’s figure to be correct, 
hardly just to themselves if they had any grounds for supposing it to be erro- 
neous, It now appears that Rumif’s figure is wonderfully reliable, for, besides his 
