22 5 
Origin and Development of the Compositce. 
approximations to the experimental values (4’5 and 2-1) obtained 
for the velocities equivalent to the critical pressures A and B. 
The minimum wind for dispersal in these three species is 
clearly approximately the same as the minimum wind required to 
blow the fruits out of the tube without a pause. It is also closely 
approximate to the value calculated from the rate of fall assuming 
the angle to be 45° and the efficiency of the pappus to he ’25. We 
have, therefore, a convenient although somewhat rough method for 
the calculation of the minimum wind required for the dispersal of 
those pappose fruits of which the rate of fall in quiet air is 
known. 
Leontopodium alpinnm, Cass.—These fruits not being pappose 
the preceding methods do not apply, and as they are spherical the 
dispersal is more analogous to spore dispersal or to the cases given 
by Thoulet (35) and Udden (37), see section D. Considering the 
very slow winds which are sufficient for the dispersal of spores it 
is probable that the velocity (4.4 m.p.h.) equivalent to the critical 
pressure A is at least an approximation to the minimum wind for 
dispersal in this species. It is interesting to note that, although 
Stokes’ Law may apply in this case, the cypsela in the Compositae 
when epappose is usually cylindrical. It has been shown by Eiffel 
(13) that a cylinder has a much slower rate of fall than a sphere 
of the same diameter, for instance, if the length of a cylinder is 
equal to the diameter and the movement is in the direction of the 
axis, the resistance of the air is five times greater than it is for a 
sphere of the same diameter moving at the same rate. 
D. Phylogenetic Significance of Fruit-Dispersal. 
The significance of the hydrodynamics of fruit dispersal may 
not he immediately obvious, but the question of whether a given 
fruit requires for its dispersal a wind of 2 m.p.h. or a wind 50 m.p.h. 
is of fundamental importance in the interpretation of the facts of 
geographical distribution. A proper understanding of the conditions 
of wind-dispersal is necessary for the rational study of the history 
of the Compositae, their migrations and colonisations, their paths 
of travel and regions of concentration, which form the subject of 
the following chapter. 
De Candolle (11), who is followed by Bentham (1,7), Praeger 
(30), Willis (IV, 94) and others, ignores or gives rather unsubstantial 
reasons for neglecting the authentic evidence for wide dispersal, and 
gives a series of negative observations in support of the dispersal for 
short distances only, That such generalisations from particular 
