CENTRAL NERVOUS SYSTEM IN VERTEBRATE ANIMALS. 
767 
bundles, one of which comes from the ventral transverse commissure, the other two 
come from the ventral horn of grey matter. I could not find a distinct connexion 
between these roots and the cells existing in the grey matter. In Mugil there are 
generally nine of these cells to be seen in one section, when any were to be seen at all. 
The Scorpcena Porcus, although a sluggish fish, has more cells in the ventral horn of 
grey matter than the Mugil, which is more active. 
The dorsal roots of the spinal nerves arise by three or more bundles, which emerge 
from the outer and dorsal side of the dorsal horn of grey matter. Some of these fibres 
came from the dorsal commissures, others from the grey matter itself. 
JSTo cells are to be seen in this horn in the Mugil, but in Scorpcena Porcus the unusual 
circumstance occurs that its spinal cord, at some distance behind the medulla oblon- 
gata, presents at its dorsal part, and close to the mid-line, a pah’ of ganglia, one 
on each side. These ganglia are formed of large cells, which resemble those only 
found on the ventral side of the cord in other fishes ; they have a distinct nucleus and 
nucleolus. 
Conclusion. 
With regard to the homologies of the brain of Teleostei, I have come to very nearly 
the same conclusion as Stieda, although my opinion being based solely on researches 
into the Teleostean nervous system may require some modifications hereafter when I 
have an opportunity of investigating the brain of Elasmobranchii and Ganoids. 
The posterior unpaired tuberosity of Teleostei I consider to correspond to the cere- 
bellum, in the ordinary acceptation of the term. The optic lobes would be the corpora 
quadrigemina, or rather bigemina, and would homologise with the anterior pair. 
Fritsch — whose work ‘ On the Minute Structure of the Brain in Fishes,’ only recently 
published, I have received since the foregoing pages were written, and to which unfor- 
tunately I have not yet had time to devote the amount of study which its importance 
demands — has opposed to this interpretation the consideration that their structure is 
too complicated ; but I think that, if the other conditions of the homologies are satis- 
fied, structure alone would not form a sufficient objection. The other conditions are 
satisfied : their position at the anterior end of the aqueduct of Sylvius, behind the third 
ventricle and in front of the cerebellum, points in that direction. The fact that they 
are comparatively highly organised simply shows that in fishes they perform functions 
which in higher animals are relegated to other organs. This being the case, the ven- 
tricle of these lobes would be simply an extremely enlarged aqueduct of Sylvius, and 
would correspond to the ventricle of the corpora quadrigemina in the foetus. The 
third ventricle is the fissure between the crura cerebri, and would extend between the 
cerebral lobes as far as the commissure which connects them, which would then homo- 
logise the anterior commissure, while the transverse bundle of fibres on the floor of 
the ventricle of the optic lobe has the relative position of the posterior commissure. 
Fritsch has marked the last-named cavity the lateral ventricle, but this appears to be 
MDCCCLXXVIII. 5 F 
