304 MR. W. GRIFFITH ON THE GENUS GNETUM. 



feriorly, terminating towards the middle of the long diameter of the albumen ; the tissue, 

 however, on the same line, and between it and the apex of the albumen being more dense, 

 and having a peculiar appearance. The walls of this cavity are rugged and irregular. To 

 the upper portion, and to one side of the cavity, is attached the embryo by means of an 

 enormously long, tortuous, and spii-ally biit irregularly-twisted cellular funiculus, the 

 cells being much elongated and twisted. Its length varies, when moderately pulled out, 

 from 3i to 5 inches, the length of the fruit being 1 inch. This funicle, as well as the 

 extremely similar one of Cijcas, has the property of contracting when immersed in water. 

 When ill situ it is tolerably closely packed ; it is dilated towards its attachment with the 

 embryo. The embryo itself is more or less ovate, its radicle tapering off superiorly, and 

 being completely continuous mth the apex of the funicle. The cotyledons are equal, very 

 small, and mutually applied by theii" plane contiguous faces. The plumule is incon- 

 spicuous, and only indicated by a rounded, entii'e, very minute elevation. The lower half 

 of the embryo is lodged in lax, apparently dislocated tissue. 



[Roxburgh * describes the seed of G. scandens as having, besides the two outer coats, 

 which he refers to the fruit, " two integuments : the exterior one thin and fibrous ; the 

 inner one less distinct, and adhering to the perisperm very firmly. Perisperm conform 

 to the seed, cartilaginous. Emlnyo in the apex of the perisperm, straight. Cotyledons 

 two, ensiform, unequal. Radicle superior." In his drawing of G. scandens the embryo is 

 represented with a short obtuse radicle, two unequal cotyledons, and an immense plumule, 

 consisting of a filiform stalk equalling in length the larger cotyledon, bearing at its 

 apex two small leaflets. Nothing is said of the funiculus.] 



From the foregoing account, I trust it will be seen that, with the exception of the 

 subsequent and sudden appearance of the additional membrane t, the development of the 

 ovula of Gnetum presents scarcely a single peculiarity worthy of notice. The changes 

 that occur in the two outer coats from an early period up to the time of the maturity of 

 the seed, I consider as of secondary importance, and as the result of modifications de- 

 pending upon theii' anomalous situation. The sudden appearance of the additional mem- 

 brane does not seem to me to be capable of any explanation : hitherto I have not met 

 with it in any stage of development prior to that when its apex reaches to the opening 

 existing in the outer coat. I can assert with tolerable confidence, that this coat does not 

 originate in a separation of the circumferential tissue of tlie nucleus, and that hence it is 

 not analogous to the tercine. I am unable to state what functions it performs, although 

 its exsertion, its free communication with the apex of the nucleus, its dilated laciniate 

 apex, and its subsequent sphacelation would lead us to suppose that it is connected with 

 impregnation. If such be really the case, it is obvious, from the period at which its 

 development is completed, that fecundation can only be effected through the agency of 

 the poUen of anthers belonging to spikes at an earlier period of growth. 



The complete agreement in development of the nucleus and albumen with that of the 



* Flora ludica, vol. iii. p. 519. 



t Roxburgh, in his figure of G. scandens, represents the exsertion of this membrane or coat as occurring during 

 tlie existence of the male flowers. 



