206 
Atkinson . — A Monograph of the 
origin from the ray-cell, and gradually approaches the wall 
toward the antherid-zone. Fig. 32 shows some of these varia- 
tions. The development of the procarp proceeds with the 
branching which forms the layers of the wall, otherwise the 
trichogyne could not penetrate through the wall, which be- 
comes quite compact by the pressure of the cells. It arises 
by the budding of a cell from the peripheral or radial face of 
the generative filament. This cell grows by elongation and 
transverse fission, forming a row of cells. Fig. 37 represents 
the procarp of six cells (subgenus Lemanea) just entering the 
space between two adjacent cells of the medullary layer, while 
the latter are producing by peripheral fission cells of the inter- 
mediate layer. One or more cells would now develop with 
the advancing development of the wall, and the trichogyne 
from the last one would pass out as the cortical cells are 
developed. 
In the species of the subgenus Lemanea the origin of the 
procarp is chiefly in the middle region of the procarp-zone, 
near the ray-cells. In Lemanea australis a great majority are 
borne on the first and second cells of the generative filaments, 
while in Lemanea annulata they are distributed farther toward 
the antherid-zone, though never so close as in the subgenus 
Sacheria. In the latter subgenus they arise mainly in the 
two ends of the procarp-zone, reaching sometimes into the 
antherid-zone, or in some species, Lemanea fluviatilis for ex- 
ample, they also arise in the middle of the procarp-zone. 
To correctly trace the origin of the procarp in the subgenus 
Sacheria a preparation should be made of a longitudinal half 
of a segment, so that the branching of the generative filaments 
can be seen. Fig. 11 represents such a preparation from 
Lemanea fucina. Nearly all of the procarps arise from the 
short branches of the generative filaments formed by the 
growth of a tie-cell into a short generative filament. Near 
the end of the generative filaments the cells become shorter 
and appear more like tie-cells in shape. A thin section by 
such a cell would make it appear that the procarp arose from 
a tie-cell. These redundant generative filaments may be so 
