14 Scientific Proceedings, Royal Dithlin Society. 



increased ; the same result was obtained in a few instances by drying 

 fruiting branches of certain species, though such cases were exceptional. 

 His conclusion is that, allowing for dried seeds and branches, about one-tenth 

 of a flora might be taken as capable of transport across a considerable 

 stretch of sea, and of subsequently germinating.' 



Other early observations on the subject are well summed up by Martins,^ 

 who himself performed a valuable series of experiments on seeds placed in 

 sea-water. He selected 98 kinds of seeds, giving preference to large 

 seeds with a thick coat, and to those of littoral plants. His results showed 

 that the majority of these seeds floated on sea-water, about one-third 

 sinking at once; also that one-third of the total were capable of germination 

 after six weeks' immersion, and one-eleventli after three months' immersion. 

 He concludes that the transport of seeds by currents plays an insignificant 

 part in the dispersal of species between countries separated by sea. 



In 1873 Thuret' published the results of several seasons' experiments on 

 the same subject. He used seeds of 251 species, belonging to seventy-seven 

 different orders. He demonstrated that the conclusion of Martins, that the 

 seeds capable of floating in sea-water were twice as numerous as those 

 incapable of the same, was incorrect as a generalization. Thuret attributed 

 the error to want of thorough wetting of the seeds experimented on ; but in 

 a foot-note to the paper, Alplionse de Candolle points out that Martins, as 

 has been stated above, selected seeds of high presumptive buoyancy — large 

 seeds with thick coats, and seeds of littoral plants. Thuret's detailed list shows 

 that more than half of his 251 species sank at once ; most of the remainder 

 had sunk at the end of one or two days ; only a very few floated for a week or 

 more. This result has been amply confirmed by subsequent observers. But 

 just as Martins selected plants which gave a buoyancy percentage too high 

 for general application, so Thuret erred on the other side. His list included 

 very few littoral species, whicli, as a group, are now known to possess an 

 especially high index of seed-buoyancy, and in consequence the 2 per 

 cent, or so, which may be deduced from his tables as representing the seed- 

 buoyancy of his plants, is too low to give an average figure for tlie seed- 

 buoyancy of a flora. 



Many detailed observations followed on the dispersal of seeds by water, 

 those of Scandinavian and Danish botanists, Kolpin Ravn* and Sernander^ 

 for instance, being especially valuable. As regards our native flora, the 



' Origin of Species, ed. 6, pp. 606-509. 

 2 Bull. Soc. Bot. de France, iv, p. 324. 1857. 



2 Bibl. Univ. et Eevue Suisse. Arch, des Sciences Phys. et Nat., N.P., xlvii, 179-194. 1873. 

 * F. Koi.PiN Eavn : Om Flydeevnen hos Froene af voreVand-og Sump-planter. Bot. Tids- 

 skvift, xix. 1894. 



' E. Sernander : Den Skandinaviska Vegetationens Spridningsbiologie. Cpsala and Berlin, 1901. 



