RELATIONSHIPS OF THE INDEFINITE INFLORESCENCES. 163 
There still remain to be considered one or two special cases. 
In the capitulum a special method may be adopted find 
accommodation for a large number of flowers by the formation 
Ss. 
As, however, the internodes remain throughout quite undeveloped, 
we need not consider this Codincéiion further here. 
appears at first sight to be a connecting link between the raceme 
and the umbel, as indeed it is stated to be by Kerner;* but if we 
apply the principles involved in the theory of tier iden, this is seen 
to be a misconception. For the corymb oe boa tinge internodes 
fully developed, and hence belongs to the e type. As will 
be seen later, this view is supported by Pass: from systematic 
We may turn now to the evidence which may be brought forward 
in support of the relationships pari aa by the theory of inter- 
nodes ; but, before doing so, I wish to lay special stress on one 
ag z 
to-day possesses a complicated inflorescence has me 
another passed through all re stages from a solitary flower 
upwards. In some cases this have ha appened to a limited 
may 
extent; but the majority of plants were Bae from ancestors, who 
were able to bequeath to their offspring a heritage of vita l energy 
sufficient to at once maintain any such complicated form as a panicle 
or an umb to the inflorescence of mene far-away primitive 
ancestor we are not here concerned. The only consideration with 
which we have to deal is the relationships of these inflorescences, 
and the causes which have led to their adoption. 
Tue Evipence ror THESE RELATIONSHIPS. 
The relationships here presented have the merit of simplicity, 
and, with one or two exceptions, the links in the chain are fairly 
obvious. The gradual increase in complexity as we pass along each 
of the two hag of development, the Racemose and the Umbellate, 
is in accordance with the axiom that complexity of structure indi- 
mn high development. If it were not for this Wats a complicated 
rm of inflorescence such as the compound umbel might be regarded 
ai the type, and all the other forms derived from it. “There i s, how- 
ever, no reason to believe that the indefinite ee tiers any 
re on to the rule, and therefore a form such as a raceme may 
ooked eon as derived from the spike. 
n-development of certain internodes in the peimaaer and 
umbel is 87 anatomical fact. Many authors, Asa Grayt among others, 
have long ago called special attention to this point. Another fact 
which has a remarkable bearing on this subject is that i : nearly all 
flowers, with a few a ae such as the Passion-flower, &c., the 
internodes remain undeveloped. - If the non- development ‘of certain 
internodes is characteristic of the flower, it is not surprising that a 
similar phenomenon should take place in the ‘floral axes of whatever 
* Kerner and Oliver, The Pcie Hidoed < Plants, i i. i. 730. 
+ Asa Gray, Structural Botany, § 246, p. 129. 
mM 2 
