IN THE CLUSfACE^, MAGNOLIACE.E, ETC. 93 



admirable figures, of the growth of the ovule recorded in the valuable Memoirs of 

 Brongniart and Mu-bcl, and, as will be seen, proved by the observations of the latter, 

 cited in the foregoing page. It is not more unreasonable to conceive that the. osseous 

 deposit, forming the crustaceous covering, may in some cases be secreted in the arilliform 

 coating, as well as in the primine of the ovule ; in both instances, the secreted matters 

 must pass through the same channel, and be supplied by the same vessels of the funicle, 

 and such depositions at one poiat, instead of another, are probably regulated by the 

 nature of the pre-existing tissues. In Zanonia and Feuillcea, the arilliform nature of the 

 outer coating is better shown by the membranaceous state of the tunic, which is extended 

 like a winged covering over the seed. 



From these circumstances we may infer, that the arilline need not necessarily be always 

 fleshy in its nature, as in Magnolia ; but that it may be either membranaceous, gela- 

 tinous, coriaceous, or even osseous in its structure. Thus I have found from the position 

 of the raphe, that the hard highly-polished tunic of the seeds of Drvmys and Ilicimn, 

 usually regarded as testa, should be held to be a true aril : thus also the coriaceous 

 coatings of many seeds wiU in Kke manner be found to be arilliform in their origin. Of 

 that kind of seminal coating where the arilline is intimately combined with the testa, and 

 where the raphe, greatly branched, lies imbedded between them, forming a comj^ound 

 tunic analogous to the structure already described in the Tovomitece, a very remarkable 

 instance occurs in the Oleacece, where the raphe, instead of being spread into numerous 

 branching nervures, exhibits itself by infinitely minute ramifications, as a dense network 

 of most delicate spiral vessels, crowded together into a cottony web, like that of a spider's 

 cocoon, and fills up the entire space between the testa and ariUine ; these tunics, aided 

 by this interposition, are closely agglutinated into an apparently simple coating, but by 

 maceration they may be sejiarated from each other, and the interposed network may be 

 drawn out into inntmierable elegant spiral threads. This structure I have found in 

 Tessarandra and Olea, and it probably exists in other genera of the family. Nearly the 

 same development occm's in Casuarina, where a thick web of spiral fibres is found inter- 

 posed between the crustaceous testa and the outer membrane, which is extended over it 

 in the form of a wing. This structure, noticed many years ago by Mr. Brown, and at 

 that period described as a singular occurrence (Gen. Rem. p. 40), has since been confirmed 

 by Schleiden, and figured in Schnitzlein's ' Iconographia ' (Gen. 86). The peUicular 

 coating of the seed, here extended in the form of a wing, which covers the excessive deve- 

 lopment of the raphe, will probably be found to be an arilline. 



"We have a good illustration of the ariUiform nature of the external coating of the seed, 

 under somewhat variable forms, in the Fassifloracece. In Tacsonia pinnatiMipitla I foimd 

 the seed invested by a mucilaginous pulpy envelope, which dries into a loose pellicular 

 vesicle, quite detached, leaving a considerable vacant space between it and the osseous 

 testa, in eveiy part save at a small point at the base, and at another in the apex : this 

 peUicle shows no vessels of any kind, except in its longitudinal i*aphe, which is imbedded 

 in its substance, appearing as a prominent white nerve, running from the basal hUum to 

 the summit, where it finds a passage through a caruncular spot (the diapyle) in the apex 

 of the testa, beneath which it becomes lost in the chalaza of the inner integument. In 



