INSULAR FLORAS. 41 



plants by this means. In this way he collected a number 

 of most interesting facts, which he supplemented by experi- 

 ments on the spot with seeds that he kept in sea-water for 

 thirty, forty or fifty days ; many of which germinated freely. 

 He observed large numbers of drift-seeds sprouting ; but 

 few of these escaped destruction by crabs. He also brought 

 away a large number of seeds, with which he has since ex- 

 perimented, and made the results known to the writer. We 

 may give one example : Seeds of Ipomcea grandiflora col- 

 lected in the Keeling Islands in September, 1888, were kept 

 dry until June, 1890, and then placed in sea-water until 

 June, 1 89 1, when they were subjected to conditions favour- 

 able to germination, which followed with vigorous after- 

 growth. One plant thus reared is still growing at Kew. 

 From time to time Mr. Guppy has sent the writer various 

 germinating seeds that had been subjected to similar con- 

 ditions. Experiments with, and observations on, the ger- 

 mination of the coco-nut, go to prove that it can rarely 

 establish itself from drift-seeds, especially where there are 

 crabs. And it should be mentioned that the instance re- 

 corded 1 of a growing coco-nut on the recently upheaved 

 Falcon Island, in the Tonga group, was not, as was at first 

 supposed, the result of accident ; for it has since been ascer- 

 tained that a chieftain of one of the neighbouring islands 

 actually planted the seeds. 



W. BOTTING HEMSLEY. 

 1 W. Botting Hemsley, in Nature, xli., p. 537. 





