and on the Structure of the Seed in the Rhamnacese. 87 



between them, except round the margin of the apical chalazal disk, 

 where they appear to be intimately agglutinated ; the embryo is 

 flat and compressed, its foliaceous cotyledons being parallel with 

 the partition of the nut, aiid their mai'gins contiguous to the 

 peripherical cord of the raphe. This organization* is more evi- 

 dently seen in this instance, as the seed is six times longer than 

 that of Colletiu, which it greatly resembles in structure : it 

 is also remarkable for exhibiting many points analogous to the 

 organization I have described in the seeds of Cucurbitaceaf . 



This singular extension and displacement of the raphe, ac- 

 companied by the formation of an extraneous crustaceous cover- 

 ing in the seed, was not altogether a novel fact, as I thought 

 when I first remarked it in the Cucurbitacece ; for the occurrence 

 was partly noticed thirty years ago, in Rhamnus, by Brongniart, 

 who, strangely enough, was not struck by the appearance of so 

 remarkable a phsenomenon, which he passed over without com- 

 ment, and it has remained unnoticed by other botanists since 

 that time. Brongniart's words are worthy of being quoted, as 

 showing a notable correspondence with my observations. Al- 

 luding to the RhamnacecB, he says J : " Dans toutes ces plantes, 

 le test lui-raerae, examine au microscope sur des ovules dejk 

 fecondes et a moitie de leur developpement, est compose de trois 

 couches tres diflferentes; Fune, externe, n'est qu'un epidermis 

 mince § ; I'autre, moyenne, est solide et fibreuse, formee de fibres 

 ou cellules alongees transversales, c'est elle qui doit former le 

 test de la graine ; enfin I'interne, tres epaisse dans les premiers 

 temps qui suivent la fecondation, est formee de parenchyme 

 lache, compose de cellules remplies de globules verts ; elle 

 s'atrophie, peu-a-peu, a mesure que I'amande et I'cmbryon se 

 developpent. C^est en general dans cette couche que passent 

 les vaisseaux nourriciers, qui composent le raphe, et vont former 

 la chalaze, raphe qui dans ce cas, suit Fun des cotes de Povule 

 en dedans du test, et redescend en partie de I'autre cote, apres 

 que la plupart des vaisseaux qui le composent, ont donnes nais- 

 sance par leur epanouissement a la chalaze." The latter conclu- 

 sion appears, however, to have been only conjectural ||. 



The entire peripherical circuit of the tracheal vessels in one 



* Drawings of this analysis mil be given in plate 33 e of the ' Con- 

 tributions to Botany.' 



t Linn. Trans, xxii. 92. % Ann. Sc. Nat. x. 340. 



§ This is a thin layer of cellular tissue that invariably lines the epicarp, 

 and partially adheres to the outer cnistaceous coating of the seed in 

 Rhamnus, and which I have not considered necessary to mention in the 

 preceding description. 



II " La chalaze vasculaire m'a toujours paru formee entierement par une 

 epanouissement des vraies trach^es : elle est formee exterieurement p.ir 

 une expansion des vaisseaux du raphe" {t. c. p. .341). 



