216 BOTANICAL GAZETTE [SEPTEMBER 
the proembryo is complete in the 8-celled stage. Thus far “pro- 
embryo” has been recognized largely as a term of convenience to 
describe the stages preceding the elongation of the suspensor. 
PoLYEMBRYONY.—The writer has found it necessary to dis- 
tinguish between the polyembryony caused by the simultaneous fer- 
tilization of several eggs and that brought about by the separation 
of the embryos of a single egg. The latter form of polyembryony, 
which is spoken of as “cleavage polyembryony,” is no doubt a 
constant feature of Pinus, and may possibly. be found in some of 
the other genera of this family. The statement that “all 4 cells 
of the lower tier may unite to form a single embryo, or they may 
separate to produce 4 embryos,’’ may hold for the Abietineae as a 
group. The writer has found separated primary embryos in all of 
the species of Pinus examined, which includes P. sylvestris, for which 
STRASBURGER reports only 1 embryo per archegonium. Forms 
like Thuja (20) seem to show splitting of the embryos at times, 
while in other cases the archegonium produces only 1 embryo. 
Coker (7) found the embryos splitting apart in Taxodium and 
also in Podocarpus (6). Some of these more modern forms are 
therefore not constant like Pinus in this respect. 
Polyembryony by cleavage from 1 egg is no doubt a primitive 
gymnosperm character, even though it has persisted to the Ephedra 
level, where it is on its way to elimination. No angiosperm has 
shown this form of polyembryony, which is a further proof that it Is 
a primitive character. Aside from its phylogenetic significance, 
the feature of polyembryony is a wonderfully effective means for 
the possible elimination of unfit embryos, involving as it does in 
Pinus some 32 embryos when 4 archegonia are fertilized. 
Although no matured twins have been found to arise by the 
cleavage of the egg in Pinus, this has been demonstrated for Ginkgo 
by Lyon (26). Here we have a close parallel to the animal twins 
which are formed by cleavage, and Lyon has shown that the 
twin embryos may originate from the same archegonium, remain. 
organically connected, and develop equally to the maturity of the 
seed 
EARLY EMBRYO IN RELATION TO OTHER CONIFER EMBRYOS.— 
The known stages of the proembryos in Picea (28), Abies (29), 
