Herrick, Morphology of Brain of Bony Fishes. 27 



tract which crosses the ansulate commissure in the region of the 

 exit of the oculomotor. 



The main body of the bundle is broken up into numerous 

 tortuous bundles which pass into and around the substance of 

 the two niduli mentioned. The whole region is filled with cells 

 of various sizes, especially the multipolar switch cells described 

 es])ecially under the account of the nidulus ruber and niger. 

 From this subthalamic region a strong multiple tract passes 

 ventro caudad into the medullary portion of the hypoaria (nidu- 

 lus niger) where the fibres enter one of the processes of the 

 enormous "switch cells" of that ganglion. A connection is 

 thus formed with the fibres of tiie ventral peduncles in an indi- 

 rect manner. This enormous development of an anastamosing 

 system may be regarded as a partial compensation for the reduc- 

 tion ot a similar system in the cerebrum. From the nidulus 

 ruber and regions immediately adjacent fibres pass to the dorsal 

 ])art of the medulla, l)ut have not been followed in detail. A 

 rather strong band passes from the dorsocephalic portion toward 

 the dorso median part of the thalamus. This tract is composed 

 of non-medullated fibres with cellular accompaniments and must 

 not be confused with the much more conspicuous tract of the 

 commissura horizontalis which, after passing from its crossing 

 caudad to the cephalo-ventral part of the nidulus, continues ob- 

 liquely through it, emerging dorsad where it meets the tract of the 

 cephalic peduncle of the cerebellum which comes into actual 

 contact with but does not enter the nidulus, then both tracts pass 

 dorso cephalad to their termini in the nidulus corticalis as else- 

 where described. The strong tract from the base of the 

 ruber was seen by Fritsch and regarded as the continuation 

 of the dorsal part of the horizontalis tract and accordingly the 

 whole was called colnmna fornicis. (Figs, i, 2, Plate VI, will 

 illustrate these points.) 



5. The ^- basal cerebral fasciculus y This tract, which was 

 described \x\ Vol. I, p. 356, is certainly crossed, and the axial 

 commissure of Edmger is accordingly a decussation. The fibres 

 originate in the region of the mesaxial lobe and decussate caudad 

 of tlie other anterior commissure bundles. They are very fine 

 and remain near the median line and part at least of the bundle 



