436 Mr. J. Miers on Ephedra. 
and becomes an embryo, when a mass of albumen is generated 
around it, which completely encases it, pushing upwards in its 
growth the persistent amnios, which, like a vitellus, remains as 
a sort of cap upon the albumen. When the seeds attain matu- 
rity, the thread of the fertile embryo, as well as the several 
filaments still bearing their sterile sacs, all having descended 
into the vitellus, there become twisted together and surrounded 
by lax cellular tissue, thus forming the suspensor. This de- 
velopment is clearly shown in its several stages Joc. cit. pl. 8. 
figs. 3, 5, 15, 16, 21, 26. After the perfection of the embryo 
within the albumen, the radicle grows upward, as if germinating, 
and, forcing a passage through the vitellus, finds an exit by the 
side of the attachment of its own suspensorial thread, and re- 
mains there naked and quite distinct from the main suspensor 
(as shown in fig. 17)—a fact claiming our especial notice, as it 
is a condition very different from what Griffiths and Gaudichaud 
show of the suspensor in Gnetum, where the amnios, having 
been altogether absorbed, the suspensor appears continuous with 
the radicular extremity of the embryo. In regard to the nature 
and origin of the several threads (which together form the sus- 
pensor), Miquel was evidently inclined towards the notion that 
they are the pollen-tubes, which had penetrated the several 
embryo-sacs, becoming afterwards agglutinated together by lax 
cellular tissue into one general twisted cord: this is certainly 
the only legitimate inference from all the circumstances recorded, 
in support of which Griffiths relates (loc. cit. p.306) that he 
always found in Cycas that “the tubular membranous portion 
of the apex of the nucleus becomes actually crammed with pol- 
len-granules, from the lower and outer of which pollen-tubes 
are pretty generally produced.” 
Changes very similar in their nature take place in the pro- 
duction and growth of the suspensor in the Conifere, the several 
stages of which were clearly described and figured by Brown 
two years prior to the publication of Miquel’s researches on 
Cycadacee ; and the singular coincidence of these developments 
is a strong argument in favour of the near affinity of these two 
families. Brown showed that, in Pinus* (which, like Cycas, is al- 
ways polyembryonous), previous to the development of the several 
embryo-sacs and their suspending threads, the former are con- 
stantly first discernible in distinct areolar cells in the nucleus, 
arranged, as in Cycas, in a circular series in the summit of the 
amniotic body: these sacs descend, each separately suspended 
by its own elongated thread; but generally only one of them 
becomes fertilized. The further progress of this growth, and the 
* Ann. Se. Nat. 2 sér. xx. p. 195, pl.5; Ann. Nat. Hist. xiii. 368, pl. 7. 
