212 TEAXSACTIONS AND PROCEEDINGS OF THE [Sess, Lxui. 



The plants which afforded material for the present 

 study (at St. Andrews) attained the large proportions 

 indicated in descriptions of the species. "When in flower 

 they form stately objects in a roomy conservatory. 



The long neck of the bulb (Fig. 19;?) is a noteworthy 

 character. On one occasion bulbs in pots were laid 

 almost horizontally under a stage in a greenhouse in 

 winter, the young leaves forming the neck being then 

 enclosed in the dry membranous basal parts of older 

 leaves. When in that position the necks curved upwards, 

 and exhibited, by reason of their massiveness, a rather 

 striking illustration of negative geotropism. 



Fertilisation of the flower is readily accomplished by 

 hand. The ovaries (Fig. lov) swell with considerable 

 rapidity. The fruit (Fig. 2) is sometimes reddish, but is 

 commonly a dull greyish green. All the examples under 

 examination were found to be solid bodies, each composed 

 of a single seed, roughly hemispherical in form, and about 

 1^ inch in diameter. The mass of the seed is composed 

 of a firm, fleshy, moist endosperm, containing a very 

 considerable number of chlorophyll granules. The large, 

 elongated, .somewhat curved embryo (Fig. 4c7») lies 

 embedded in the endosperm. 



If seeds be left for a few weeks, even on so dry 

 a place as a shelf, they will germinate. The radicle 

 makes its appearance as a stout peg (Fig. 3r), but at no 

 determinate point of the seed. It is stated by Herbert^ 

 that the germination of Crlnnni seeds can be expedited 

 " by cutting carefully off a portion of the fleshy mass, so 

 as to expose the point of the embryo "; and, further, that 

 "the operation requires a cautious liaiul, for if the point 

 is cut by the knife, the vitality of the seed is 

 destroyed." 



In tlie first experiment, a seed was kept in a desk until 

 the cotyledonary process, terminated by the radicle, had 

 protruded I inch (Fig. 3), The seed was then dissected, 

 so as to expose the embryo longitudinally, but still leave it 

 adherent by one side to the endosperm (Fig. 4). The 

 part cut off was replaced, and the whole then bound 

 together and laid lialf-sunk in the soil, the radicle being 

 1 Ilerhert, AiniirylliilacoEe: p. 402. 



