CONTRIBUTIONS TO BOTANY. 
67 
leaves in the primary axil only are simple, in the second and 
third axils they are deeply bifid, but in all the following nodes 
they are, as usual, 3-foIiolate, or cirrhosely conjugated : in the 
simple leaf just mentioned, its petiole is as long as the petiole 
and petiolule conjoined of the conjugate leaf, showing that this 
circumstance arises from its stipuloid character, or is owing 
solely to the suppression of one of the folioles from imperfect 
growth, or its decadence in the early weak state of the plant, and 
is not the complete development ; it cannot therefore be adduced 
as an exception to the ordinary rule. 
There ai’e a few species, among EuhignoniecE, with heterophyl- 
lous leaves, that oflFer an exception to the general rule ; these, 
however, are not properly climbers, but are of an erect and 
short stature j Bignonia brachypoda, DC., represents the type, 
and among them may be classed three plants hitherto placed in 
Cuspidaria. Most of the instances I have seen of these truly 
heterophyllous species are generically alike, and they consti- 
tute a group that may be regarded as a subgenus of Arra- 
hidea. By this isolation, and with this exception, we maintain 
a constancy in the rule of foliaceous development that I have 
advocated. In the group just mentioned w^e meet with hetero- 
morphous varieties, where, in consequence of some morpholo- 
gical change in a few of their many 3-foliolate leaves, two 
or all three leaflets grow together, assuming the state of a 
single leaf, or of an unequal pair of conjugate leaves of an 
unusual gibbous form : the mode of distribution of the nerves 
in . such cases shows that such coalescence is due to the same 
kind of monstrous growth which, under similar exceptional 
circumstances, we see in other families. As it sometimes hap- 
pens that the cirrhus is wanting, it may be urged that one of 
the folioles of a conjugate leaf may also be suppressed, and the 
other reduced to the state of a simple leaf : of the possibility of 
this occurrence there can be no doubt ; but observation shows 
that such instances are extremely rare, and then not universal 
in the same plant, and must be held to be entirely of an excep- 
tional character. Under the genus Panterpa, of which B. leuco- 
pogon, Cham., is the type, I have explained how, and under what 
circumstances, simple and compound leaves occasionally occur in 
the same plant ; but it is there shown that the simple leaves in 
such cases partake of a stipuloid nature. These instances there- 
fore cannot be said to affect the general rule above indicated. 
There is sometimes a peculiarity in the ligneous structure of 
the BignoniacecB that merits attention ; the stems of many of 
them, in their transverse section, exhibit strong medullary rays, 
not radiating from one common central point, as generally occurs, 
but disposed in parallel decussating plates, in the form of a cross, 
K 2 
