154 REV. PROF. G. HENSLOW ON THE ORIGIN OF 
Besides the last three genera of Polygonaceze, a few others might be mentiond, as 
genera of Lythrariez, Kc. 
The question is, how did these ternary verticils arise, if the ordinary phyllotaxis is 2° 
With the exception of Calycanthacee, which have opposite leaves, all the rest have, as a 
rule, the quincuncial arrangement, or 2. 
An important fact to observe is the ease with which the phyllotaxis can be changed 
from one angular divergence to another, or from any one kind to what I have elsewhere 
called “ tricussate ” whorls, i. e., successive whorls of threes alternating in position. This 
is facilitated by the fact that there are always three leaves, or whatever may represent 
them, in any portion of the spiral of any divergence (excepting, of course, 4), “cut at 
random " and sufficient to make a circle when projected on a plane. 
As the first illustration we may take Chimonanthus *. This has an indefinite number 
of bracts, sepals, and petals, with no break between them except in the size and colouring 
of the innermost members. The diagram given by Le Maout and Decaisne + shows the 
5 arrangement in the outer series concluding internally with eight larger parts, which 
may be regarded as petals, and arranged on the 3 plan. Then follow 5 perfect stamens 
and 5 +5 abortive ones. In the centre is represented a whorl of 5+5 carpels; thus the 
$ divergence prevails throughout, with one departure in the petals. 
A point here worth noticing—which is well illustrated by Chimonanthee as well as by 
members of Cactaceee—is that the calyx is developed out of leaves, or rather bracts; 
while the corolla—though, as a rule, it issues out of stamens—can be produced as well 
from “below upwards,” instead of the reverse, as in Water-Lilies. In Calyeanthus 
there are no breaks between bracts, sepals, and petals, the abrupt change occurs with 
the stamens. The coloration increases inwards as well as the size of the parts. In 
Epiphyllum the colour is the same throughout. 
. That the coloured di-trimerous whorls of the calyx of Anemone issue from the leaves 
is indicated by the fact that one or more leaves of the involucre have not infrequently the 
same form and colour as the true sepals; while, on the other hand, the outer stamens 
may become more or less petaloid staminodia and secrete honey (4. Pulsatilla), as in 
Atragene compared with Clematis. 
Moreover, when polyandrous flowers become double, the calyx usually remains 
unaltered, while the number of petals may increase to any extent. 
When, therefore, the sepals increase in number, they are added to from below, as in 
Anemone japonica, in which it may consist of 3, 3+3, or 3+3+3 whorls; while 
in A. Hepatica the involucre becomes a floral whorl. . 
I wil now explain how di-trimerous whorls—i. e, two whorls of three parts 
each—correspond to ove cycle of the 2 arrangement and require wo complete coils 
or circles. 
Figure 1 (p. 155) represents the 2 divergence; numbers 1 to 5 form the first cycle $, 
but the éwo coils are not completed until No. 6, over No. 1, is reached. This—which, 
* This genus and Calycanthus are the only ones of the order Calyeanthacew. 
T ‘ Descriptive and Analytical Botany,’ p. 191. 
tf Thi gives rise to the usual quinary arrangement in Dicotyledonous flowers. 
