68 MESSRS. NEWELL ARBER AND J. PARKIN ON 
In the case of the Bennettiteze, however, the Mesozoic descendants of this 
group, we find both the male and female organs aggregated into an amphi- 
sporangiate strobilus, and further that the megasporophylls are of a highly 
advanced type, and have undergone great reduction, as well as possibly other 
extreme modifications. The stage reached in the evolution of the micro- 
sporophylls is obviously greatly behind that of the megasporophylls. They 
show hardly any marked advance beyond the condition of affairs met with in 
the Pteridospermesm. The microsporophylls are still essentially compound, 
fertile fronds. Any progress in evolution is confined to the synangium, which 
is still the dominant type of male fructification. and perhaps more highly 
evolved in the Bennettite than in any known Pteridosperms. The stamen, 
per se, is quite a recent innovation, so far as this line of descent 1s concerned. 
But the adoption of entomophily, by means of closed carpels, which in the 
ultimate analysis will, we believe, be found to be the real influence which 
called the Angiosperms into being, no doubt involved considerable modifica- 
tion in other parts of the flower, and among these the male organs. The 
incoming of this type of pollination, thereby effecting an immense saving 
in the amount of pollen produetion necessary to ensure cross-fertilization 
(see p. 74), seems to have been the signal for considerable reduction in the 
male fronds of the pro-anthostrobilus. Eventually a much simpler structure 
has been evolved, consisting of a sporangiophore bearing two synangia. 
Although we regard the microsporophyll of the Angiosperms as derived 
originally from a highly branched organ, by reduction, there would seem to 
he very few cases among living members of the group in which a survival of 
this ancient feature can be traced. It is just possible that such may occur 
among the Myrtaceæ, e. g., Calothamnus, and possibly also in Ricinus, where 
the stamens are pinnately branched, but in the Polypetalous orders, such as 
Capparidacese, Dilleniaceze, ВезеЧасе», Hypericacese, Cistaceze, Malvacere, &e., 
in which so-called branched or divided stamens are found, this phenomenon 
is of a different nature *, and has no direct bearing on this discussion. On 
the other hand, it is admitted that the gap between the male organs of the 
Bennettiteze and the Angiosperms i is a big one, and that we are not at present 
able to trace the various stages in the reduetion of the mier osporophyll. 
The Perianth. 
The Bennettitean cone possesses а basal, spirally arranged series of sterile, 
leaf-like organs, which form an integral part of the strobilus. We imagine 
that the pro-anthostrobilus of the Hemiangiospermez also possessed this 
feature, which we interpret as an undifferentiated, primitive perianth. With 
the assumption of entomophily, and the consequent evolution of the Angio- 
* (Goebel (1905) p. 535. 
