THE BULBIFORM SEEDS OF CERTAIN AMARYLLIDE^ 371 



plete account of these structures was given by a former Secretary of 

 our [the Royal Horticultural] Society, Richard Anthony Salisbury. 

 Salisbury's great desire was to publish a (Tcmera I'lantamm, but the 

 work never appeared. At his death in 1829 he left a large quantity 

 of MSS. and beautifully executed drawings, which are now in the 

 Department of Botany at the British Museum. A fragment of the 

 Genera was printed in 1806 ; it comprises a considerable portion of 

 the petaloid monocotyledons. Salisbury subdivides AmanjUidacm, 

 as we now understand them, into a number of orders, one of which, 

 AmanjlUdeie- (p. 120), is distinguished from all the others by cha- 

 racters of stamens and corolla, and, "what I deem most essential, 

 in the bulbiform fleshy seeds, hitherto accompanied with a solid 

 peduncle ; so that when we cannot obtain the former, a tolerably 

 good conjecture of their nature may be formed by the latter. These 

 bulbiform seeds are often whitish or tinged with pink till exposed 

 to the air, when they gradually assume a green hue, sometimes so 

 dark as to be nearly black, but howsoever dark they may be always 

 known by their thick fleshy coat, hitherto in AmanjUide(B devoid of 

 albumen ; if only a few in each cell, they are generally large ^and 

 irregularly shaped, not unhke small Potatos." He criticizes Ker's 

 suggestion as to their being an accidental and alternate mode of 

 fructification, and says: "After a great many enquiries of our 

 nurserymen and gardeners, I do not hesitate to reply, that all 

 those species whicli have these bulbiform seeds never produce any 

 other sort; neither are they pecuhar to Auiayyllidea;, but occur in 

 the preceding as well as the following orders of Pancratem and 

 Strumarece; here, however, they begin and terminate for aught 

 I know to the contrary." "With respect to their structure," he 

 says, "many which I first dissected in 1790 at different periods of 

 their growth, from the distinct vessels near their margin left no 

 doubt in my mind that the great mass consisted of a thick fleshy 

 coat." He also criticizes Brown's statement that in some cases the 

 seed separates before the embryo is formed; "many observations, 

 lately repeated out of deference to his authority, convince me that 

 these bulbiform seeds, so far from being detached before their 

 embryo becomes visible, adhere to the dissepiments of the peri- 

 carpium till it is not only formed but very often sprouts." The 

 radicular edge of the embryo "is invariably directed towards the 

 micropyle, but when the seed swells to a large size this is removed 

 by dilation of the hilum to a considerable distance from the 

 nourishing duct, being placed at the opposite end of the hilum 

 as in Leiiuminoaa ; and by the time many of these seeds are ripe, 

 all traces both of micropyle and hilum, except the cicatrix of the 

 nourishing ducts, nearly vanish ; the original disc of the hilum is, 

 however, often concave. After the radicle comes out of the fleshy 

 coat at the micropyle, the facility with which it forces a passage 

 through other substances is astonishing, rarely turnhig out of its 

 way, but piercing an adjacent seed of the opposite cell in those 



• Comprises Crinum, Ammocharis, Buphane, Amaryllis, Brunsvigia, Nerine, 

 Lycoris, Hei<>;ea, and Carpolyza. 



