52 TlMEHRI. 
bility is that the two varieties of roots have similar 
functions." The experiments however show that the root 
development of the cane is very great ; they show also 
that when cane is planted deeply under ordinary circum- 
stances more time is necessary for the shoot to appear 
above ground ; but that in the meanwhile the roots are 
not idle, for the fleshy ones, which in all probability form 
the mainstay of the future stool, develop very freely. 
