SAND OR IC U M KOETJAPE AND DENDEORIUM CANINUM 
173 
tlie wood is lisrht red. The two forms were confused by Blume, 
by Miquel, and by Koorders and Valeton. According to Dr. Smith, 
Sandoricum indicum Blume is ketjapi and S. nervosum Blume is 
sentoel, with the native names interchanged. Thus S. nervosum 
Blume Bijdr. 1G3 is a synonym of S. indicum Cav. = $. koetjape 
(Burin, f.) Men*., and is not the same as S. Maingayi Hiern as 
Mr. Ridley thinks. I am not concerned with the question as to 
whether one or two species exist here. Dr. Smith thinks that the 
sentoel and ketjapi cannot be maintained as species. Is it not 
possible that the ketjapi is a form derived from the sentoel by culti¬ 
vation and selection ? At any rate the application of the specific 
name koetjape would seem to be fixed. 
In reference to Dendrobium caninum (Burin, f.) Men*., Mr. Rid¬ 
ley is correct in maintaining that this binomial does not appertain 
to the Pigeon Orchid, Dendrobium crumenatum Sw. Burman’s 
descriptive sentence reads: “ caulescens, foliis ovatis aveniis inte- 
the 
) he 
gives as a reference to his description “ Angrecum caninum Humph. 
Amb. 6, p. 105, t. 47. f. 1 ” ; nothing in his original descriptive 
sentence is quoted from Rumphius. The descriptive sentence and the 
local name “Angrecutan. Javanis ” indicate clearly that Burman 
had an actual Javan specimen. Dr. Hochreutiner informs me that 
he could not locate Burman’s type of Fpidendrum caninum in the 
Delessert Herbarium, but that Dendrobium crumenatum Sw., is 
represented as Fpidendrum spatulatum (FI. Ind. p. 188, angrec 
poeti) ; for Burman at least, the Pigeon Orchid, D. crumenatum Sw. 
was F. spatulatum , although perhaps not F. spatulatum Linn., which 
is supposed to be a Vanda. The figure cited by Burman in the 
original description of Dpidendrum caninum represents Dendrobium 
anosmum Lindl. (D. superbum Reiclib. f.). If we follow Ridley’s 
interpretation and consider the Rumphian illustration as the type, 
then the binomial Dendrobium caninum (Bunn, f.) Merr. would of 
necessity have to be accepted as the earliest name for the species 
commonly known as D. superbum Reichb. f. Reichenbaeh’s species, 
however, does not occur in Java, and hence it is very improbable that 
Burman’s type, whatever it may have been, represents either Reichen¬ 
baeh’s species or the form figured by Rumphius. Under the circum¬ 
stances it would seem that the binomial Dendrobium caninum must 
be abandoned. 
If Mr. Ridley were correct in maintaining that Dendrobium 
crumenatum Sw. was based on Acjrcecum caninum Rumph., this 
specific name would have to be restricted to the Amboina form— 
Dendrobium pupilioniferum J. J. Sm.—and some other name adopted 
for the widely distributed Malayan one. Fortunately in this case 
Swartz’s species was actually based on a Javan specimen, as examina¬ 
tion of the original description clearly shows, and is not typified by 
the Rumphian reference as Ridley claims. 
1 would emphasize the fact that in Burman’s Flora Indica as 
well as in Loureiro’s Flora Cocliincliinensis there is little evidence 
that any of the binomials therein contained were primarily based on 
gerrimis, nectario conico.” Ridley states that “Burman quotes 
figure and description from Rumphius.” This is true only in tha 
