202 
James Small. 
Cumberland. Such material is, however, pabulum for the philo¬ 
logist and archaeologist rather than the botanist. 
Dispersal by Water. 
Among the first observations of fruit dispersal in the Com¬ 
posite© is a series of experiments by Hoffmann (24) on the floating 
powers of a number of species, most of which sank within twenty 
days. De Candolle (11) and Hemsley (23) give rivers and ocean 
currents as agents which disperse fruits to great distances, but no 
case of such dispersal in the Composite is quoted, although 
dispersal on floating logs and icebergs, as mentioned by Darwin 
(II, 16) and Guppy (22), is quite possible. Kronfeld (VII, 3) showed 
that pappus fruits such as those of the dandelion floated much 
longer with the pappus attached, even when it is closed. The 
closed pappus usually surrounds an air bubble which assists in the 
flotation. 
In the case of Krakatau (15) some of the species of Wedelia 
and possibly also those of Conyza and Blumea were transported by 
ocean currents, and in the somewhat similar case of Taal Volcano 
(19-20) Wedelia biflora and Eclipta alba most probably arrived by 
water. The latest addition to the meagre list of water-dispersed 
Compositae is Ambrosia crithmifolia, which is given by Guppy (22) 
as transported on drifting logs. 
Dispersal by Wind. 
The pappus has usually been regarded as an efficient mechanism 
for dispersal by wind, and there have been many general observa¬ 
tions of the structure on this account, such as those of Hildebrand 
(V, 30-32), who also mentions cases (V, 30) of wind dispersal of the 
whole capitulum where the involucral bracts act as wings. Kronfeld 
(VII, 3) considered that in the case of Tragopogon pratensis the 
lateral cilia of the setae produce a structure which acts as if it were 
a continuous membrane. Taliew (V, 65), Steinbrinck (V, 64), 
Hirsch (V, 33), Prieb (18) and others have noted the biological 
significance of the hygroscopic movements of the setae and involucral 
bracts, dispersal being favoured by dry winds and prevented by 
moist winds. 
Poerste (17) mentions a curious addition to the dispersal 
mechanism in Ambrosia trifida. The achenes are beaked, with five 
short protuberances at the base, and five or six strands of ice have 
been observed attached to these on frosty mornings: as these 
“ ice-curls” were 1^ to 3 inches long Poerste suggests that they aid 
