3?o 
Stephens . — The Embryo-Sac and EmbVyo of 
dermatogen is a single external layer. It may be anticipated that a root- 
cap is formed, when germination begins, by the periclinal division of the 
dermatogen overlying the tip, but as uniform failure has attended all attempts 
to germinate the seeds, this point is still in doubt. 
The chief peculiarities of the embryo are thus : — 
(1) Absence of a suspensor. 
(2) The great development of the hypocotyl and its adaptation thus 
early in life to the function of food storage. 
(3) The reduction in size of the cotyledons. 
(4) The absence of a root-cap at the stage found in the mature seed. 
Discussion. 
Before discussing the nature of this type of embryo-sac it may be worth 
while to note a few facts regarding the affinities and distribution of the 
little-known order in which it occurs. The Penaeaceae show no very close 
relationships to other orders, except to the monotypic Geissolomaceae, 
which is united to them by Bentham and Hooker. Both Engler and 
Bentham and Hooker have placed the order near the Thymeleaceae 
(Engler among the Thymeleales, Bentham and Hooker among the Daph- 
nales), and this seems to be its natural position ; Van Tieghem, however, 
mainly on the strength of its anatomical characters, places it next to the 
Melastomaceae. 1 It was hoped that this investigation might throw some 
light upon its true systematic position, but the whole life-history here dis- 
closed is so abnormal that it merely serves to isolate the order still further. 
It is true that Modilewski has recently described in Euphorbia procera 2 an 
embryo-sac which, so far as his investigations have been carried (from the 
four-nucleate stage to the first divisions of the embryo), shows a structure 
almost exactly the same as that here described. But as the embryo-sac in 
other species of Euphorbia is of the normal type, and as the two orders 
otherwise show no points of resemblance sufficient to warrant a close 
relationship, there appears to be no doubt that this is a case of parallel 
development. As regards the other orders with which the Penaeaceae are 
grouped, the Thymeleaceae is the only one of whose life-history any account 
has been published, Strasburger 3 having recently investigated species of 
Wiksstroemia , Daphne , and Gnidia , none of which show a similar depar- 
ture from the normal. The same may be said for Cryptadenia uniflora , 4 
of which I have made a preliminary examination. Geissoloma marginata , 
the sole representative of the Geissolomaceae, appears to show other 
peculiarities, which will be described later. 
The peculiarly limited distribution of this order may be of interest 
in connexion with its rather isolated systematic position. It is entirely 
1 Van Tieghem, 1893, p. 291. 2 Modilewski, 1909. 3 Strasburger, 1909. 
4 Mr. A. J. Ballantine kindly preserved material of this for me. 
