OF THE RAPHE IN SEEDS. _ 103 



found the end of the radicle extending- a little beyond the mouth of the cavity : this con- 

 jecture is confii-med by the drawings of Gaertner (De Fruct. tab. 208). A similar per- 

 sistence of the embryo-sac, and its forming the lining of a large cavity in the albumen, 

 which is thus nearly separated into two parallel plates, and the occasional protrusion of 

 this sac beyond the limit of the albumen, occur also in Stri/chuos and Ljnatla, as may be 

 seen from the figures given by Gaertner (DeEruct. tab. 189). The protrusion of a portion 

 of the embryo-sac, and with it of the radicle, in the manner above conjectured to exist in 

 Diospi/ros, is known to occur- in Myzodendron, where it is avcU figured by Dr. Hooker in 

 liis ' Flora Antarctica,' plate 101. fig. 18 & 19; and in plate 105. fig. 17, 18 & 19*. A 

 singular example of the persistence of the embryo-sac is afforded in MantHtacece : this 

 occurs in the genus Thalia, where two large vacuities are seen in its copious albmnen, on 

 each side of the embryo : these are the persistent sacs in which two abortive embryos had 

 existed : this curious fact was fh'st recorded by Mr. Brown nearly fifty years ago (Prodr. 

 p. 307) : it is one of the few instances that occm*, where several embryo-sacs are generated 

 in the same ovule. 



In regard to the origin of the alliumen in seeds, it was shown by Gaertner {loc. cit. 

 Introd. 138), and Mr. Brown confii-ms it as a well-established fact (Linn. Trans, x. 36), 

 that it is merely that condensed portion of the amniotic fluid which remains unabsorbed 

 by the embryo at the termination of its growth. This excessive portion of the fluid is 

 generally expelled from the embryo-sac (quintine, Mirbel), and is deposited and consoli- 

 dated in the cellular tissue of the quartine, which lines the tercine (nucleus, Broton) when 

 it constitutes the albumen, and in ordinary cases the sac becomes absorbed without 

 leaving any trace of its existence. Besides the instances of its permanence just given, its 

 existence under other peculiar circumstances was first shown by Mr. Bro^vn, in the 

 NymphcBacece, where a portion of the redundant amniotic fluid forms a separate and 

 second albuminous deposit, immediately surrounding the embryo, which he has aptly 

 named the vitellus, and which in the ripe seed is found lodged in an exserted portion of 

 the embryo-sac : it is therefore always seen at one extremity, outside of the great mass of 

 the ordinary albumen : when this occurs, it is accomjoanied by the appearance of a long 

 tubular vacant space in the axis of the albumen which is the opposite extremity of the 

 persistent embryo-sac, of which the other portion containing the vitellus is a continua- 

 tion. This has been shown to exist among Dicotyledones, in the Nymphceacea, Cabom- 

 bacece, Saurnracea;, Fiperacea;, &c., and in several cases among Monocotyledoiies (Appen- 

 dix, King's Voy. p. 21). A very excellent illustration of this occurrence may be seen m 

 the admii-able work of Dr. Asa Gray (Gen. PL Un. St.), where evidence of its presence is 

 shown in the seeds of Nymplicea, Niiphar, Cabomba, and Bmsenia, in pi. 38, 39, 13 & 44. 

 Mr. Brown observes (Joe. cit. p. 22), " that the albumen, properly so called, may be formed 

 by a deposition or secretion of granular matter in the utriculi of the amnios " (embryo- 

 sac) " or in those of the nucleus itself" (tercine or quartine) ; "or lastly, that these two 



* Schleiden (Linncea, xi. 52G. tab. 1 1 . figs. 12, 13) has sliown the curious fact, that in Ceratophyllum the principal 

 portion of the embryo is developed outside of the embryo-sac, which in the ripe seed remains investing the plumule, 

 and enclosed within the cotyledons ; a small quantity of albumen is deposited in the embryo-sac around the plumule, 

 but none is formed outside of the large cotyledons, which are invested only by a simple thin integumentary covering. 



