436 
and why, in the other form of phyllotaxy, the effort should take 
place in relays till the first column is again reached. ne 
Considering the number of primitive leaves in the seed, it is 
obvious that we must regard the normal phyllotaxy of exogens as 
originally whorled and with two leaves in a whorl, while the leaves 
of endogens have, at first,aspiral arrangement. This conception, 
however, must not be taken as a practical guide for the study of 
mature plants, as it would prove misleading in a large proportion 
of cases. It isalso obvious that the flower is, in nearly every 1n- 
stance, a compact series of whorls, whatever the arrangement of 
the foliage leaves. 
The whorled phyllotaxy is very simply subdivided, according 
to the number of leaves ina circle. This number, omitting the 
primitive leaves and the flowers, is the same for every whorl of 
any individual, although it may differ in individuals of the same 
species. For example, I have found growing within a few feet of 
one another, Lupazoria with five, six and seven leaves, respectively, 
in a whorl; yet all were of the same species (Z. purpureum, I be- 
lieve). and, from their proximity, may very probably have grown 
from seeds of the same plant. Like almost every botanical fea- 
ture not connected with the flower, phyllotaxy is a very unreliable 
guide to generic and family grouping. Thus, all the mints have 
opposite leaves, but the closely-allied family of figworts includes 
examples of both general kinds of phyllotaxy; none of the genera 
of the composite family embrace species illustrating both kinds, 
but the genus Cornus includes both alternate-leaved and opposite- 
leaved dog-woods. 
My notes on phyllotaxy as a guide to analysis, relate almost 
without exception to the trees of North-eastern America, and are 
supplementary to the excellent table of Charles S. Newhall. All 
subsequent statements, must, therefore, be understood with this 
qualification. 
First of all, is to be noted a great preponderance of the spiral 
phyllotaxy, there being only seven genera represented, with Opp? 
site leaves—Acer, Negundo, Fraxinus, AEsculus, Viburnum, Chito- 
nanthus, and the single species, florida, of Cornus. Catalpa is the =) 
_ only instance of a tree with more than two leaves in a whorl, hav- a 
_ ing usually three (but sometimes only two.—Wood). 
