376 John F. Holm 



Commissura anterior. In reptiles and in Cyprinus carpio Edinger (9) 

 mentions a nerve tract corresponding to the one I described as connect- 

 ing both lobes, to which tract he gives the name of Ramus connectens 

 lob. olf. In the reptilian anterior commissure there are fibres crossing 

 from the cortical olfactory centres and Epistriatum as well as the 

 Ramus connectens. In Myxine the cortical tracts of the Commissura 

 anterior are naturally absent, but I have also been unable to find 

 signs of any tracts connecting the Epistriata either in stained or 

 GoLGi preparations. It is, however, very possible that a connection 

 may exist between the striata or epistriata through a commissure 

 situated somewhat above and caudally to the Chiasma nervorum 

 opticorum and probably homologous to the Commissura postoptica 

 superior described by Haller (17) in Scyllium. Part of the fibres 

 passing through this commissure spread in a fanlike manner, most of 

 them running in a dorsal and caudal direction towards the Mes- 

 encephalon, but some passing along towards the striata or Epistriata 

 regions. Besides these fibres, this commissure includes those that 

 spring from two ganglia situated one each side of the medial line in 

 the Hypothalamic region. These Ganglia hypothalamica consist of 

 small nerve cells, and in Golgi preparations it is easy to follow the 

 fibres from one ganglion to the other. The Striatum region is con- 

 nected with the Thalamus by a nerve tract, the Tractus strio-thala- 

 micus, that always exists in other vertebrates. This tract is not very 

 well defined in Myxine, but can be seen as a diffuse fascicle emerg- 

 ing from the lower rostral region of the brain and running along 

 towards the cell nuclei of the Thalamus (Fig. 16). 



The cell-nuclei of the Thalamencephalon 



are difficult to homologise with the nuclei in higher animals, but 

 one of them is well defined and is in an analogous position to the 

 Nucleus rotundus in other fishes. It is situated somewhat anteriorly 

 and laterally to the Ganglia habenulae and descends far down into 

 the brain (Fig. 16). This nucleus consists of bi- or tripolar cells 

 (Fig. 15) a little larger than the cells generally found in the brain. 

 The Nucleus rotundus seems to be connected with different parts 

 of the brain. It receives some fibres from the Tractus strio-thala- 

 micus and some from the Gang, or Commissura habenulae. With 

 the Golgi method I have found in transverse sections fibres running 

 from this ganglion which at first pass upwards, then turn laterally 



