40 BOTANICAL GAZETTE - [juny 
meeting at the apex. The change in its contents by which 
it is also made conspicuous in the later stages will be described 
later. 
One of the peripheral cells first produced, which has already 
cut off an upper sterile cell, now grows rich in protoplasm and 
becomes the supporting cell of the carpogenic branch, cutting 
off the basal cell at one side, but apparently not always on the 
same side, as may be seen by comparing figs. § and g. In fig. 
§ two of the peripheral cells have cut off each a sterile cell. 
The third peripheral cell has been severed from its mother cell, 
but the point of former contact may still be distinguished on 
each cell. In fig. 9 the mother cell of the carpogenic branch 
has divided into two cells, which are still in very close connec- 
tion. Here the supporting cell of the carpogenic branch is the 
only one with an ‘“‘appendicular” cell, and as neither of the 
others shows indications that one has been broken off, it is prob- 
able that no other has been produced. 
In none of the mature carpogenic branches observed were — 
fewer than four cells found. The branch is curved so as to form 
almost a right angle at the second cell (figs. ro-13). The. 
terminal cell, the carpogonium, is smaller than the others, nearly 
triangular in outline, and prolonged into a trichogyne from the 
pointed apex. 
The trichogyne is an elongated slender structure, as long as 
the remainder of the fruiting branch, and is straight or only 
slightly curved, lying in a direction parallel with the axis of the 
branch. Its diameter is slightly greater at the apex than in the 
portion adjoining the carpogonium. Its contents do not take 
a fuchsin stain, but a decidedly granular appearance is observ- 
able. 
In fig. 11 a peculiar development is shown, and a somewhat 
similar, though not quite identical, appearance was observed in 
another case. The supporting cell of the carpogenic branch 
has given rise, not to a single sterile cell, but to a branched 
cellular filament. The carpogonium is slightly larger than usual, 
and no trace of a trichogyne is seen. It seems probable that in 
