Origin and Development of the Composite. 227 
experimental results here given, nor with the theoretical view which 
enables us to calculate the minimum wind for dispersal, nor with 
Praeger’s observation of the path of the thistle down, nor with 
Udden’s observations on the path of solid particles in an 8 m.p.h. 
wind. The close accordance of the experimental velocities with 
the velocities as calculated from the observed rate of fall shows 
that the relation of the rate of fall to the velocity of the wind 
required forsuspension of the fruit for an indefinite period is at least 
approximately that discussed in Section C. 
The relative humidity of the air, however, has a very marked 
effect on the efficiency of the pappus as a dispersal mechanism, 
and the study of the meteorological conditions obtaining in the 
region of dispersal becomes necessary. In Britain, France, and 
Ceylon, which are the dispersal regions considered by the above 
mentioned opponents of longdistance dispersals, wind is associated 
with a high relative humidity and a low relative humidity is 
associated with calm. The conditions necessary for long distance 
dispersal are a low relative humidity combined with a wind always 
above the minimun value ofW for the particular fruit and blowing 
in the same direction over a large stretch of land or water. These 
conditions are comparatively rare in northern Europe and in Ceylon, 
but occur occasionally in most of the regions in which longdistance 
dispersal has been observed. 
As a result of the above investigation the writer has reached 
the conclusion that given these three conditions there is no 
limit to the distance of the dispersal of pappose fruits, and with 
this conclusion Dr. J. W. Evans, from his experience of the 
dispersal of sand, is in complete agreement. The occurrence of 
pappose species with a limited distribution is quite in accordance 
with the Law of Age and Area which will be discussed in the next 
chapter. The comparison by Bentham (I, 7) of the local species of 
Senecio with wide spread species of epappose genera is an example 
of the special pleading which has been used to explain away the 
obvious fact that the pappus-mechanism is an efficient means of 
dispersal. De Candolle’s comparison by percentages of pappose 
and epappose species in the various tribes (11) is just as 
superficial. Before such comparisons (or one of widespread pappose 
species with local epappose species) can be used every factor in the 
two cases must be balanced, These factors would include age, 
other means of dispersal, available habitats, the environment in 
which the journey starts and ends, colonising powers, physiological 
