Origin and Development of the Composite. 203 
the wind dispersal of the fruits. Another curious but better known 
anomaly is the dispersal mechanism of Parthenium hyslerophorus. 
Here the bracts enclosing two male florets are attached to the 
top of the achene of each female floret, and these bracts with 
the withered male florets enclosed act as wings. 
The mechanics of the pappus have been considered by 
Dingier (12), Praeger (30), Mattei (28) and Dandeno (9), who all 
regard the mechanism as a simple parachute. 
Apart from these general observations there is the controversial 
question of the distance to which a pappose fruit can be dispersed 
by the wind. De Candolle (11), although he regarded the wind as 
“ la cause la plus gdnerale et la plus ordinaire de dissemination 
des especes sur toute la surface d’un pays,” maintained that there 
was no evidence for the transportation of seeds over even narrow 
arms of the sea. He was, moreover, sceptical about the records of 
grains of sand, etc., being blown for long distances by the wind. 
Bentham (I, 7) also considered that a few miles was the limit for 
the dispersal of pappose fruits even by strong winds. Kerner (25), 
and more recently Beauverd (1), have shown that in the Alps 
pappose fruits are raised to a considerable height, but they both 
consider that all such fruits fall when the upward current fails in 
the evening, and that they come to rest very close to their starting 
point. 
Praeger (30), with whom Guppy (22) agrees, calculates from 
the rate of fall in quiet air that many pappose fruits would 
require an initial elevation of about one mile before they could 
be blown to a distance of fifty miles by a wind of 50 m.p.h. Heavier 
fruits would require a still greater elevation according to Praeger. 
Willis, although he states (IV, 94) that the Orchids and Composite 
are best suited for long distance dispersal, and considers that wind- 
carried seeds arrive more frequently than other Angiosperms on 
hill-tops, maintains as a result of his observations on hill-top floras 
(43-45) that such cases of long distance dispersal are rare even in 
the Composite, and that “we have little evidence to show that it 
occurs for instance between one continent and another.” 
In spite of the objection of lack of evidence for long distance 
dispersal which is raised repeatedly by the above-mentioned 
authors, the positive evidence for such dispersal makes a longer 
list, and has secured more adherents than the view of the opposi¬ 
tion. Wallace, for instance, in his “ Island Life ” says, “we are as 
