24 Nr. 4. L. KoLpErup ROSENVINGE: 
nucleus, for a transfer of a fertile nucleus from the 
carpogonium cannot have taken place. The cell bore two 
carpogonial branches but neither of them had produced a 
normal carpogonium; the end-cells were filled with proto- 
plasm in both branches, and one of them showed a distinct 
nucleus. A similar stage is shown in fig. 10 C, where the 
auxiliary cell also seems to contain several nuclei and no 
carpogonium is developed. In fig. 12 D, the bearing cell 
has taken an angular shape, but fig. 12 E shows more ad- 
vanced stages; three bearing or auxiliary cells are here 
seen, one, very similar to that in fig. 10 D, having an 
irregular rectangular shape and bearing a two-celled car- 
pogonial branch toward the lower surface of the leaflet, 
and two others, much larger, pushing out several long 
protuberances in all directions, partly penetrating between 
the surrounding cells of the foliole. The cell to the left 
contains several nuclei partly entering the protuberances. 
The carpogonial branches of these two procarps seem to 
be more or less degenerated. The next stage is represented 
in fig. 13 B where the auxiliary cell has become still more 
enlarged and the protuberances have produced, at their end, 
cells connected with the large cell by long threads of 
protoplasma. These cells form branched rows forcing their 
way between the surrounding cells of the gametophyte. A 
good deal of the parallel filaments above in the figure, 
forming a low excrescence on the mother organ probably 
derive from the auxiliary cell, but in the present case 
it was not possible to distinguish such cells from those of 
the gametophyte. The cells of the carpogonial branch could 
no longer be recognized. A similar or a slightly more ad- 
vanced stage is shown in the Plate fig. I, where the large 
cell is seen near the centre of a young nemathecial wart. 
