48 Journal of Comparative Neurology. 



In sections of the bass it has proven possible to trace the 

 connection between the cells of the ectal ganglionic zone and 

 ventral part of the optic fibre layer satisfactorily, while in most 

 cases it is only possible to determine that the ventral process, 

 which is very strong, suddenly turns out of the plane of the sec- 

 tion. In those cases the fibres of the optic tract also fall out of 

 the plane. 



The connections of the other fibre tracts must be left for 

 future study. 



In Amphibia we should naturally expect to find the struc- 

 ture of the tectum most simply developed. A glance at the fig- 

 ures on Plate X will show that we have not been deceived. 

 Fig. 1 2 is a section through the ental portion of the tectum as 

 seen with a one-twelfth oil immersion lense, drawn with the cam- 

 era to a scale. Some of the cells are drawn in from adjacent 

 portions of the field, i. e. the section has been somewhat 

 crowded, otherwise it is accurate. The nuclei of the epithelium 

 or ependyma are large and more or less quadrangular, staining 

 more deeply than the nerve cells above. In nearly all of the 

 latter there may be demonstrated a long process from the spar- 

 ingly developed protoplasm extending peripherad and dividing 

 into a number of fine branches. The horizontal cells in the 

 tract near the middle of the tectum have been found in only a 

 few cases and seem to be cells like those beneath, which have 

 been drawn out of their normal position by the growth of the 

 tract. The greater number of the fibres cf the horizontal tract 

 seen in the drawing can be be followed caudad to the root of the 

 fifth nerve. They arise in processes of the large cells shown at 

 a, Fig. 5, of the same plate and are doubtless homologous with 

 the giant cells of the mesencephalic nidulus of the trigeminal 

 nerve seen in all classes of vertebrates. The fibres from the 

 horizontal cells above referred to pass cephalad and seem ulti- 

 mately to curve peripherad and enter a meshwork produced by the 

 ental ganglion cells. That these horizontally placed cells are 

 homologous with the ental nerve cells is rendered probable by 

 the fact that at the meso-caudal end of the tectum such cells are 

 numerous (Fig. ii) and are obviously related with the ental 

 series. In a few cases two processes were seen passing, one to 

 the right and the other to the left side of the optic lobe. In this 



