3 
: ts 
Sometimes they are connate, or adnate to the sides ei val veel 
of the inner corona, up to the point of origin of the la ’ 
form a 
r seem 
can be relied upon as definitely separating the anne ey» 3 
to blend and ‘telieince vin ra slain that in many cases defies 
classification. t in 
I believe this intermingling of characters has been eb eioptir 
this way: the Asclepiadacee all require the i eg Rite se rate distance 
about fertilisation ; and two species growing within a “Ailiaed with each 
of each other would become very liable to be shimaroa Is h would not 
other’s pollen, and hybrids would be likely to as =e the seeds, 
necessarily be found in the neighbourhood of their Moin tie diablo tbe 
being provided with a large tuft of long fine hairs, wou ait: Wre 
carried to a considerable distance by the wind, in ee in another 
thistle seeds are carried, and the new hybrid established to other 
lace, where it in turn, in course of time, might aah ine species 
ybrids. That such has been the origin of many 
