410 THE TORUS LONGITUDINALIS OF THE TELEOST BRAIN: 



branches which enter the ventral portion of the cord and probably run to the muscu- 

 lature (see Sargent, : 04) . 



Sala describes and shows in his figures a third process coming off from the torus 

 cells, of greater diameter than the others. He found it took an irregular course and 

 terminated freely in a slight enlargement. It is usually directed towards the median 

 plane. He makes no attempt to explain the relations or functions of this process. 

 After studying his figures and description it is quite evident to me, from the manner of 

 its ending, and failure to establish connections with other nerve elements, that this is 

 an incompletely impregnated fibre. I think there can be little doubt that this is 

 the partly impregnated axon of the cell, which runs caudad and cephalad to emerge 

 into the ventricle, either in the ventral fissure of the torus or anterior to the posterior 

 commissure. Unfortunately all his figures are of transverse sections, and apparently 

 from the posterior part of the torus. In such sections it would be very difficult to 

 determine the true course and relations of these axons even if they were completely 

 impregnated. 



V. FUNCTION. 



We have seen that the cells of the torus longitudinalis are in connection by their 

 different processes with the endings of the optic nerve, and probably by another 

 process, with the cerebellum. They send their axons into the ventricle to form by 

 their union the so-called fibre of Reissner. Elsewhere I have shown (Sargent, :01 a , 

 :04) that this constitutes a nerve path between the optic centres and the muscula- 

 ture, and acts as a short circuit for the transmission of motor impulses arising from 

 optic stimuli. The torus, then, is the nerve-centre for the receipt of those impulses 

 coming in over the optic nerves, which call for quick reflexes. 



VI. HOMOLOGY. 



It follows, then, that the cells of the torus which give rise to Reissner's fibre 

 constitute a nidulus of cells of common function, and are homologous with the cells 

 of the same function which occur in the anterior, dorsal portion of the optic lobes 

 of other vertebrates, and have been variously designated as the "Dachkern" and 

 "nucleus magnocellularis," and which I have fully described in another paper as the 

 cells of the optic reflex apparatus (Sargent, : 04) . 



The cells of the torus of teleosts are small, while the homologous cells in other 

 groups of vertebrates are conspicuous for their great size. Why there is this marked 



