40 MESSRS. NEWELL ARBER AND J. PARKIN ON 
like Lacistema, with a complete perianth whorl. It is much more natural to 
suppose that the other members of the whorl have been aborted, and that such 
genera as Piper and Peperomia have been evolved by further reduction, 
during which the perianth has disappeared altogether. In the monotypie 
genus Lactoris, placed by Engler in the Ranales, but by Bentham and Hooker, 
among others, in the Рірегасеғ, сап be found а synthetic type linking these 
two cohorts together *. 
Thus, in our opinion, the more obvious and plausible view is that the 
Piperales branched off, probably at an early period, from the Ranales, and, as 
in the case of many other Angiosperms, have suffered considerable reduction 
in the individual flower, so much so that in many instances they have lost their 
perianth. This line of evolution appears to have progressed side by side witha 
tendency to aggregate the flowers into dense spikes ; the bracts assuming more 
and more the functions originally performed by the perianth. Ina few of the 
Piperales the grouping of the flowers has advanced a stage further. In the 
Peppers of the section Potomorpha the spikes are arranged in umbels. Such 
compound inflorescences surely are hardly characteristic of “plants of low 
organisation? f. 
Both the recent studies of the seed-development of the Piperales by 
Johnson f, and of the seedling-structure by T. G. Hill$, have led these 
authors to conclude that this is not to be regarded as a primitive cohort. 
Amentifera. 
The term Amentiferz is used here, for the sake of convenience, to include 
those families of trees characterized by unisexual flowers—or at least the 
male flowers—crowded together into very dense and definite inflorescences 
known as catkins; a type of inflorescence which is shed entire, and thus 
functions largely as a single flower. In some families, e. g., those included in 
Engler’s cohort Fagales, the catkin is of a highly complex and compressed 
nature—a feature hardly suggestive of primitiveness, but rather of a reduction 
in the component parts. In such, a suppression of the perianth might 
naturally be expected. . 
Salicacee.—This family consists of two genera only, Salix and Populus, 
the one entomophilous, the other anemophilous. Divergeney of opinion 
exists as to whether the entomophilous habit of Saliz—an almost unique 
occurrence amongst the Amentiferz:e — is. to be regarded as a primitive 
feature, or one derived from a Poplar-like ancestor by a change in the method 
of pollination. 
Chamberlain |, from an embryological study, concluded that the genus 
* Hallier (1901?). t Willis (1904) p. 515. { Johnson (1905). 
$ Hill, T. G. (1906). | Chamberlain (1897). 
