pRAEGER — The Buoyancij of the Seeds of some Britannic Plants. 17 



species of llaphauus, it may be a portion of the fruit, with a seed enclosed. 

 In the case of fleshy fruits, several conditions of dispersal are possible. A 

 fresh berry may fall directly into a stream ; or it may fall and get dried, 

 and subsequently become immersed; or it may be eaten by a bird, and 

 the wet seeds dropped into water, or dropped on a dry place and carried 

 into a stream after dessication. So far as material served, buoyancy was 

 tested under several or all of these conditions. 



Where material was obtainable, two or even three batches of seed of 

 one species from different sources were tested. Sometimes the results 

 obtained from different batches of seed were uniform ; occasionally, they 

 differed widely. To this point I shall return presently. Where more 

 than one figure is given in the table opposite the name of a species, 

 whether in the same column or in different columns, it represents the 

 flotation-periods of the most buoyant seeds of different batches. 



Explanation of the Table. 



In order that the varying buoyancy of the seeds of different species 

 may be more readily discerned, the results obtained are arranged in five 

 columns in the table which follows : — The cross in the first column shows 

 the species whose seeds sink at once ; the second column gives (in iiours) 

 periods up to one day ; the third gives (in days) periods up to one week ; 

 the fourth (in weeks) up to one month ; the fifth gives (in months) periods 

 over one month, up to 15 months, when observations ceased. The seeds 

 which were still floating at the conclusion of the months' observations are 

 shown thus — 15 +. 



In a sixth column are added flotation-periods as given by Guppy. 

 Gruppy's tabulated results are less detailed than my own, and a full 

 comparison between the two cannot therefore be made. The form in 

 which his observations are given, the symbols which he employs, aud the 

 symbols by which his results are shown ia the sixth column of my table, 

 are as follows : — 



Guppy's symbols, it will be seen, are comparable only to those given 

 in the fourth and fifth columns of my table; lesults corresponding to 

 those of my first three columns being shown by a blank in his table. 



