THE REV. С HENSLOW ON THE ORIGIN OF FLORAL JESTIVATIONS. 181 
Or again, if we start with a quincuncial arrangement, the equitant condition is 
arrived at by suppressing either the first or fifth member. This, I think, has probably 
been the case with Actea (p. 186) and Adora * (р. 474), as well as with the Crucifere t. 
Circea (Tab. XXV. fig. 2), however, I am inclined to think owes its equitant sestivation 
either to an original dimerous symmetry derived from opposite leaves, or else it may be 
due to “symmetrical reduction" from a tetramerous type which prevails in the same 
order (Onagracec) to which it belongs. 
If the equitant sstivation is presumably derived from а quincuncial whorl of five 
parts, by the arrest of one, the four remaining will probably be wholly equitant at first ; 
but they may undergo further changes. If it be the result of a convolute estivation, 
then the remaining pairs will more likely be half-equitant. Thus, if Cireewa be derived 
from а tetramerous type with convolute petals, as occurs in Epilobium, by the arrest of 
two opposite parts, then it will be seen how each part of the remaining pair will overlap 
the other reciprocally, and the resulting zestivation will become half-equitant (Tab. XXV. 
fig. 2). 
I would not, however, venture to lay very much stress on this supposed difference in 
the origin of half-equitant as compared with equitant whorls; for the overlapping of 
another by the edge of any part may be simply due to some slight advantage in growth 
quite irrespective of any cause arising from antecedent conditions—as, indeed, may be 
the ease with the calyx of Papaver, which is half-equitant, though, as I presume, its 
dimerous character is derived from a tertiary type. 
2. Tristichous.—This will be seen to be identical with ап imbricate-proper restivation 
when the whorl consists of three members only. It is very common in the perianth- 
leaves of Monocotyledons, occurring in the outer whorl about 40 p. c., and in the inner 
whorl about 60 p. c., as deduced from the diagrams of Le Maout and Decaisne. This 
possibly would require some modification ; for I found the convolute æstivation also very 
common, as in about one third of the flowers of Crocus examined. This kind is repre- 
sented by those authors as occurring in only four genera. Іп а pound of dates I found 
the percentage of the inner whorl of the perianth as tristichous to be 82, and as con- 
volute 18. Тһе outer whorl is gamophyllous, and reveals no succession of parts. 
The origin of the prevailing ternary arrangement in Monocotyledons I take to be 
simply due to the fact that it does not arise out of opposite leaves, which latter, as I 
have shown elsewhere, gives rise chiefly to the pentastichous arrangement (Trans. Linn. 
бос. sec. ser. vol. i. р. 37), so common іп Dicotyledons; hence, as the usual phyllo- 
tactical requirements demand three leaves for every projected circle, and the floral 
requirements demand an alternation of position in the successive whorls, so by these two 
effects there issues the usual alternation of groups of threes. 2; 
In Dicotyledons there is a group of orders with а ternary arrangement prevailing 
amongst them (Magnoliacee, Anonacee, Menispermacee, Berberidacee), associated 
with a few genera or orders having a large number of imbricated parts in their fioral 
whorls (Dilleniacee, Calycanthacee, Nympheacee, &c.). When the whorls are in 
* See Note B, р. 194, on the variations of the symmetry of the flowers of Adowa. 
t See Note А, p. 191, on the origin of the arrangement ofthe parts of a cruciferous flower. 
