Page:Proceedings of the Royal Society of London Vol 69.djvu/510

490 than the right, owing apparently to the great development of the right ganglion habenulse, which projects into the brain-cavity and more or less blocks up the angle between roof and side wall. Posteriorly (fig. 6) they appear to terminate as grooves at about

FIG. 6.

the level of the hinder margin of the posterior commissure, but a band of long columnar cells appears to be continued backwards for a short distance after their groove-like character has disappeared, so that we find such a band on either side, bordering the narrow opening between the iter below and the cavity of the choroid plexus of the mid-brain above.

The most striking difference as compared with the New Zealand Ammoccete concerns the arrangement of the choroid plexus of the mid-brain, which no longer dips down into the iter in the form of a deep, continuous, vertical septum (compare figs. 3 and 6). (Such a septum is, however, simulated in the sections by a mass of granular material containing what look like nuclei irregularly scattered through it. This appearance, which is not represented in the figures, suggests that the septum has undergone degeneration, or possibly the granular mass is simply a coagulum containing the remains of cells shed from