16 Scientific Proceedings, Royal Duhlin Society. 



Thoroughly dry seed was used in preference to seed just taken from a 

 fresh, moist seed-vessel; though the latter is the condition in which 

 the majority of seeds commence their dispersal-adventures, the former 

 condition must frequently occur in nature. Seed gathered fresh and stored 

 in a dry room for some months was used in preference to any other. Fleshy 

 fruits were tested in a thoroughly dried as well as in a fresh condition. 

 Drying has often an important effect in the case of fleshy fruits, while in 

 the case of hard seeds its effect is usually inappreciable. 



The experiments of Guppy' showed that between fresh and salt water 

 a very slight difference exists as regards their effect on the buoyancy of seeds. 

 After a number of tests to satisfy myself of this, I fell back on the more 

 easily obtainable medium, and used fresh water throughout the experiments, 

 resorting to salt water occasionally as a cheek. There is hardly an exception 

 to the rule that seeds which sink in the one medium sink also in the other. 

 The only effect of the salt water is to slightly increase the period of flotation ; 

 and since, as stated above, care was taken to obtain a maximum period 

 by using dried seed (and also, as appears below, by taking the period of 

 the most efficient seed of each batch as representing its buoyancy-period), 

 I believe my figures are already quite as high as we have any right to accept 

 as a buoyancy-index even for sea-water. 



The seeds, twenty to a hundred in number, wlienever so great a 

 number was available, were first cursorily examined for soundness, and 

 then thoroughly shaken up with water in test-tubes, care being taken to 

 remove adherent air-bubbles. The minority which did not sink were kept, 

 and examined and shaken up twice a day, and transferred to fresli water 

 occasionally until all the seeds — excepting occasionally a few whose soundness 

 there was reason to doubt — had sunk. This maximum period was then 

 eutered opposite the name of the species in a copy of the " London 

 Catalogue." Thus, for those seeds which did not all sink witliin a minute 

 or so, 12 hours is the unit of time, and signifies anything up to 12 hours ; 

 and so on for greater periods. Tlie maximum efliciency is thus not alone 

 entered, but in many cases somewhat exaggerated; however, the exaggeration 

 is unimportant except for short periods, and even then has little significance, 

 as seeds of short flotation-periods are ineffective for oversea dispersal. 



After a flotation-period of a month or so, the seeds were examined and 

 shaken up only once a day, and after about three months only once a week. 



In the case of a considerable number of species, the dispersal-unit may 

 be either the whole fruit or part of it. In the majority of cases it is the 

 seed itself. In many others it is a dry iudehiscent fruit. In cases like the 



1 he. cit., p. 89. 



