288 
Thus it is shown thatin Achirus the cells do not extend caudad 
of the 13th spinal nerve. The cells are smaller than in any other 
form examined, 0,055 mm in diameter, and are also further dorsad 
in the median fissure, being in some cases actually outside of the sur- 
face of the cord. 
The dendrites are very numerous and large and are gathered into 
bunches on those surfaces of the cell which are near other cells. They 
interlock with the dendrites of those cells to which they are near and 
with the best lenses and material I was unable to decide that they did 
not anastomose. If it was not a physiological continuity, it was at 
least a physical union that occurs between these dendrites. 
The neurites are very small and weak. They apparently pass to 
the same fibre bundle that the corresponding neurites of Para- 
lichthys and Pleuronectes do, but, although this was reasonably 
well seen, it was not determined beyond possible doubt. 
Paralichthys oblongus. Bothus maculatus. 
These two species were examined but no important differences 
could be detected between the apparatus as found in them and as 
found in Paralichthys dentatus. 
The Giant Ganglion Cells as found in the Larval State. 
Bothus maculatus. 
The first larval forms examined were surface-swimming flat-fish 
of from 2 to 5!/, mm in length and with the right eyes in all stages 
of rotation. The specimens were dextral (lying on the right side) and 
highly pigmented with black. From what could be learnt concerning 
them, they were probably the surface-swimming larvae of Bothus 
maculatus. 
In the smaller specimens of from 2 to 3 mm the eye was not yet 
rotated, but all traces of the yolk-sac had disappeared. The neuron 
was tube-like and composed of a wall of embryonic nerve cells, some 
of which had but the faintest traces of the ganglion cell form, while 
in the brain indirect cell division was going on actively. At the an- 
terior end of the dorsal wall of the primitive cord the first giant cells 
had appeared. They were large and crowded together in a double row 
(see Fig. 3). As this row was followed caudad the cells became smaller 
on account of being in an earlier stage of development, until, at the 
16th segment, the cells of this row had become so small that they 
could not be distinguished from the surrounding embryonic nerve cells. 
92 giant cells could be looked upon as such and these were contained 
