Dr. B. Scemann on the Bignoniacce. 197 
“ My dear Sir, “ Tlammersmith, Feb, 13, 1862, 
“JT return your paper with many thanks for its perusal, and for 
your courtesy in sending it prior to its publication. You are per- 
fectly justified in maintaining your former convictions in regard to 
Tanaecium, if you still believe in them ; but I can hardly conceive, in 
the present state of science, how it is possible for any one to conclude 
that the T'anaecium Jaroba, Sw., and 7’. parasiticum, Sw., with such 
diametrically opposite characters, can belong to the same genus, or 
even to the same tribe. Putting aside for the moment the question 
of the fruit, we find that the structure of the ovary, seated ona 
peculiar disk, in the former, is quite that of 4denocalymna and of a 
few congeners, while that of Sch/egelia has its ovules fixed in the 
middle of the dissepiment. In regard to the fruit which I described 
as that of 7’. aldiflorum, it is true that it is'not accompanied by any 
flowering specimen; but, coupled with the fact of the structure of 
the ovary, which I have fully verified, the evidence becomes almost 
complete ; for the fruit in question agrees in size with Swartz’s de- 
scription in its singular oistuew shape, its hard, smooth, 2-locular, 
2-valved shell, with ‘‘ many large, broad, compressed, imbricated 
seeds’’——characters that scarcely leave a doubt as to its specific iden- 
tity. This, again, is confirmed by the coincidence of flowering spe- 
cimens of 7’. prelongum and fruit, both sent from British Guiana 
by Schomburgk*. The structure of the ovary, about which a doubt 
cannot be raised, shows the true position of T'anaecium, and proves 
incontestably that it cannot belong to Crescentiaceze. 
‘With regard to Parmentiera, I regretted that you had not given 
more tangible characters of its fruit, and had not shown the struc- 
ture of the ovary. I referred, in the absence of these, to your draw- 
ing, which marks, by two very distinct transversal lines, that the 
fruit is 2-valvular, no such sutural lines being found in Crescentia. 
I argued therefore that 7/ these (ycur own) indications be confirmed, 
and if the ovary be found to be 2-locular, with ovules peculiarly 
placed, then Parmentiera ought to be referred to Bignoniaceze. This 
would of course include Catalpese, where it would go under certain 
conditions to be proved. I think you will not venture to gainsay so 
legitimate an inference. The characters to which you seem to 
attach so much importance—of flowers issuing from the trunk (also 
partial in other families), and of the edible fruit—are of no value in 
an ordinal point of view, whatever consideration they may deserve 
as generic attributest. They would seem to show a close affinity 
between Colea and Parmentiera. 
*“What I mentioned about Colea was founded on the statements 
recorded up to that time by the best authorities; if those facts be 
erroneous according to the evidence you have since obtained at the 
Mauritius, you must deal fairly with the inferences previously drawn 
* The fruit from British Guiana in the British Museum here alluded to 
is not accompanied by any herbarium specimens, though it is quite true 
that Schomburgk did send a Tanaecium in flower from that locality.—B. 8. 
+ I did not say they possessed any ordinal value, but simply quoted 
them in proof of the fruit being fleshy and heavier than a mere dry cap- 
sule.—B. 8S. 
