Albuca corymbosa and juncifolia to Insect Fertilisation. 369 



also larger than that of fertilised flowers. The seeds in 

 both capsules are good, only three in each failing to ger- 

 minate. In Series III. the number of unfertilised flowers, 

 as compared with the fertilised, is very large, being 7 to 2, 

 and from the larger capsule fifteen out of seventy-three 

 seeds do not germinate. This series bears most signi- 

 ficantly on the question of the cleistogamy of the flowers. 

 Series IV, might have been expected to give other results. 

 It is in Series V., as contrasted with Series III., that the 

 crucial test lies. Of 12 flowers, fecundated with pollen 

 from the functional stamens of distinct plants, only 2 are 

 unfertilised, whereas where self-fecundation is resorted to, 

 7 out of 9 are unfertilised. Again, comparing the fruit in 

 the two series, that of the former is of more uniform and 

 greater excellence. The two capsules wanting in that 

 series were good. One was sent to Mr Lindsay, curator of 

 the Eoyal Botanic G-arden, Edinburgh, and the other to 

 Kew. Before being sent, 51 seeds were taken from the 

 latter, and of them 49 germinated. The high vitality of 

 the seeds in Series V. is obvious, the average number of 

 ungerminated seeds in each of the eight pots being only 3 '6. 



The conclusions to be drawn from the experiments are 

 briefly as follow : — (1) that the pollen borne by the half- 

 aborted anthers may sometimes be potent, either to fer- 

 tilise the same flower (?) or the flowers of a distinct plant ; 

 (2) that the pollen of the truly functional stamen of one 

 flower may often be impotent as regards its operations on 

 another flower of the same plant; (3) that self-impregnated 

 flowers are almost invariably sterile ; (4) that cross-fecunda- 

 tion almost always results in fertility. 



A severe blow is dealt at the theory of the cleistogamy 

 of this Albuca, by the fact that out of 97 flowers the 63 

 unimpregnated artificially were not fertilised. If self- 

 fertilisation were the natural mode, surely out of this large 

 number a few would have borne fruit. 



After the completion of the experiments with Albuca 

 corymbosa, a single plant of Albuca juncifolia, Baker,* 

 came into flower. The latter difi'ers conspicuously from 

 the former in having cernuous flowers. The central (axial) 

 projection of the stigma was invariably present. The tip 

 of it projects so far out as to form a knob in the bottom of 



* Bot. Mag., t. 6395; Gardeners' Chronicle, 1876, p. 534. 



