428 BOTANICAL GAZETTE [ JUNE 
an axillary bud may be developed from leaf /4, instead of from 
leaf 75. Also, a bifurcation is not uncommon, and this depends 
upon the non-development of the large bud, 6’, which is then 
replaced by two other buds in the axils of leaves /* and /5, both 
of which are then situated to the right and left of the axis. 
These two buds develop simultaneously into two horizontal 
branches, with elongated internodes. Furthermore a bud may 
be developed in the axil of the leaf 7*, but remain dormant. 
When the large bud, 4*, beccmes injured, two different cases 
were observed in which the rhizome nevertheless became able 
to continue its growth. Either a bud developed from leaf /*, to 
grow out and replace the nor- 
mal continuation of the rhi- 
zome, or the terminal bud, B&B, 
ceased to be dormant, develop- 
ing into a horizontal branch. 
This last case is figured in the 
accompanying drawing (jfig.9). gEZ 
There are three relatively short | 
internodes with scale leaves, oe 
terminated by a large bud, Fic. 9. Rhizome of P. peltatum; the 
which contains a floral axis and terminal bud has grown out into a horizon- 
some vegetative buds. The sg haiaeen Micra zee 
age of this branch is three years. The first internodes show 4 
tendency to become stretched like those of the sympodial shoot 
when it develops. That the terminal bud became developed in 
this case was merely due to the injury of the shoot 6’, as is to 
be seen in fig. g. A still more peculiar case may occur when 
the bud, 4', does not grow out as a long branch, but as a few 
very short internodes, simulating the one described above. 
The development of these buds is, therefore, not restricted to 
any of the scale-like leaves, excepting the leaf, /*; neither 's 
their position constant, although the majority of cases examined 
show that a frontal position is the most common and naturally 
the most advantageous to the plant. ? 
It now remains to mention the position of the floral axis. 
