294. DR. P. GROOM ON THE LONGITUDINAL 
is the terminal compound inflorescence—and the leaves begin to show signs of decussate 
or peculiar paired arrangements. 
PHYTOLACCACE.E. 
The Phytolaccacez apparently include only alternate-leaved forms. Despite of the 
fixed nature of the acyclic phyllotaxis the internode-curves are very irregular, and there 
is no doubt that analysis into two subcurves removes the irregularities entirely, or at 
least lessens them. 
Analysis of the extremely irregular internode-curve into two regular subcurves may 
be noted in connexion with the main axis of Phytolacca esculenta I. (Curve 21) and the 
branches of P. decandra I. (Curve 231), P. esculenta II. (Curve 221). In the first 
branches of these two the depression in the middle of the displacement-subcurves is 
doubtless due to this eurve being transitional between a normal (ascending-descending) 
one and a purely descending one. 
Analysis of the extremely irregular internode-curves into two considerably less 
irregular subeurves is represented by the relatively main stem of Phytolacca polyandra 
(Curve 24) and Branch A of Ercilla spicata (Curve 25 1). 
Thus the internode-curves of Phytolaccaceze point clearly towards the view that the 
phyllotaxis of the ancestor was cyclic. In harmony with this stands the fact that 
at the base of the main stems of the species of Phytolacca examined the divergences of 
the leaves more or less perfectly conform with those of decussate phyllotaxis. But, as 
pointed out in my previous paper, this might represent a necessary mode of transition 
