94 



JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



development is going on [i.e. the base of the cotyledonary sheath- 

 surrounding the phmiule] swells and takes on a bulb form." 



In 1840 a Dutch botanist, H. C. van Hall (" Tijdsch. Nat. Geschied." 

 vii. pp. 140-164), gave a full and well-illustrated account of the fruit, seed' 

 and method of germination in Crinum capense. He takes the same ^iew 

 of the structure of the bulbiform seed as did Richard (see p. 146) ; his- 

 figures (fig. 23) show well the elongation of the cotyledonary sheath carrying 

 doTNTiwards the small radicle, the upper end of the cotyledon remaining in 

 the seed to form a swollen sucker by means of which the nourishment in^ 

 the endosperm is gradually absorbed. The plumule is surrounded by the 

 base of the cotyledonary sheath, where the bulb very soon begins to- 

 develop, the sheath forming the outermost scale. His figures also illustrate 

 the difierent length which the cotyledonary sheath attains under difterent 

 circumstances. In one case where a seed was allowed to germinate on 

 the edge of a board, and not supplied either with food or moisture, the 

 radicle was carried vertically downwards by a cotyledonary growth six 

 times the largest diameter of the seed in length, and still showed nO' 

 trace of the leaf succeeding the cotyledon. 



Later workers enable us to reconcile the differing statements as to the 

 exact nature of the fleshy mass surrounding the embryo. In 1857 Henry 

 Baillon ("'Bull. Soc. Bot. Fr." iv. p. 1020) showed that in HymenocalUs 

 speciosa the two integuments of the ovule after fertilisation become much 

 thickened and fuse together with the nucellus to form the thick fleshv mass 

 surrounding the embryo. Vascular tissue derived from the outer integu- 

 ment can be seen. In the next year Prillieux (''Ann. Sci. Nat." ser. 4, 

 ix. (1858), 97) confirmed Baillon's statements on HijmenocnUis (except that 

 he states that the fleshy coat arises purely from the primine), but showed 

 that in Amaryllis Belladonna, Crinum erubescens, C. giganteum, C. 

 taitense and C. capense the ovules are naked, and that the fleshy coat is 

 derived from a large development of endosperm, on the outside of which 

 the remains of the nucellus forms a thin membrane. Moreover, no 

 vascular tissue occurs in the fleshy coat. 



A. Braun (" Ann. Sci. Nat." ser. 4, xiv. (1860), p. 9) shortly afterwards 

 confirmed Prillieux's observations on the occurrence of two kinds of fleshy 

 seeds, which he named bulbous, where the outer of the two integuments of 

 the ovule forms the fleshy seed-coat (as in Hi/menocallis), a,nd tuberculous 

 (as in Crinum, &c.) respectively. He also drew attention to the fact, noted 

 by Bro^\Ti, that in some of the fleshy seeds (those in which their separation 

 precedes the visible formation of the embryo) spiral vessels do occiu* in 

 the fleshy mass, though Brown had previously stated in the " Prodromus 

 (p. 297) that the mass was purely cellular. The recognition of the 

 existence of the two kinds of seeds helped to explain these differences. 

 Braun also noted that several embryos might occur in one seed in 

 HymenocalUs. 



A third kind of bulbiform seed was subsequently described by Baillon in 

 an allied genus, Calostemma, in the Proceedings of the Association 

 Francaise " (Lyons, 1873). Calostemma was one of the Australian 

 <]renera to which Brown referred in his orisfinal note in the " Prodi'omus." 

 In C. Cunninghami each of the three ovary-chambers contained two 

 anatropous ovules, the development of which Baillon found to be at first 



