52 Scientific Proceedings^ Eor/al Diihlin Society. 



the fleshy-fruited genera {Prunus, Ruhus, Rosa, Crataegus, Pijrus) will 

 be referred to in a subsequent section. 



Onagraceae. 

 Species tested, 8. Most Epilobiums float for up to 1 day. 



Cramulaceae. 

 Of 7 Sedums, 8. Telephium alone had any floating power (up to 2 days). 

 Cotyledon Umbilicus floated for a few hours. 



Saxifragaceae. 

 Species tested, 8. Of 6 Saxifrages, S. stellaris alone can float (1 day). 

 Parnassia palustris floats for several weeks. 



Umbel/i/erae. 



Species tested, 34. Very variable. The species with apparently indefinite 

 floating power are Hydrocoti/le vulgaris, Oenantlte aqiiatica, Angelica syhcstris- 

 Two maritime plants, Haloscias scoticum and Crithmum maritimum, follow 

 with 2| and 5 months respectively. Quppy adds Sium latifolium, 8. erectum, 

 Peueedanum palustre (all 1-6 months) and Oenanthe crocata (6-12 months). 

 The 16 species which sink at once or within a day are all mesophile or 

 xerophile (excliiding halophile). 



Ruhiaceae. 



Species tested, 9. Variable. The only marsh-plant among them, Galium 

 palustre, far exceeds the others in buoyancy. 



Valerianaceae. 

 The hollow chambers in the fruits of Valerianella do not much assist 

 buoyancy, the maxima for four species being 2f , 21-, 5, and 14 days. 



ConqMsitae. 

 Species tested, 73. While only a few (16) sink at once, the only ones 

 which float for over a week are Artemisia maritima, Bidens cernua, B. tripartita, 

 Carduus pratensis, Helmintliia echioides, the last being the only one which 

 appeared to possess indefinite floating power. As regards the two species of 

 Bidens, Gupp}' gives them each a period of 6-12 months; but my cernua 

 sank within 3 weeks, and tripartita within 5 days. All through the 

 Oompositae, periods of | to 5 days prevail. The pappus-bearing species as a 

 group show no greater nor less buoyancy than those whose fruits are devoid 

 of appendages. In this connexion Moriz Kronfeld,' experimenting with 

 Taraxacum and Crepis foetida, found that the buoyancy of the fruits was 



' tjber einige Verbreitungsmittel der Compositenfruchte. Sitz. k. Akad der Wissensch., AVien, 

 Math.-Nat. Klasse, xci, Abth. i, pp. 414-428. 1885. 



