370 John F. Holm 



by the release of pressure wlieu the animals were taken from the 

 deep. The Fossa rhomboidal! s varies in form to a certain extent 

 and where it is most developed will be found to project deep down 

 towards the base of the brain 'Fig. 4 pi. XIX). It sends one canal in 

 rosti'al direction towards the Gang, haben, and one upwards which 

 afterwards divides into two and enters the Tectum opticum of the 

 Mesencephalon. In some individuals the canal first mentioned, which 

 is generally considered homologous with Aquaeductus Sylvii, is not 

 entirely obliterated but may be followed below the Gang, haben. 

 (Fig. 1 G.h)y where, after widening into a small cavity (Fig. 2), the 

 canal bends into right angle towards the base of the brain. After 

 about two thirds of the distance towards the base another canal 

 leaves this one and leads in straight line and rostal direction to 

 the Trigonum cinereum, which is thereby connected with the ven- 

 tricular system. This last canal is not uncommon and it has existed 

 in at least half the brains investigated. I call this canal >Canalis 

 connectens« and the vertical which gradually widens towards the 

 base of the brain and enters the infundibular cavity »Canalis ventri- 

 culi terti«. This cavity has been described by G. Retzius (39) who 

 was the first to discover those peculiar lateral canals which lead 

 horizontally from the infundibular cavity into the base of the brain. 

 A horizontal section on the level of these canals through the base of 

 the brain shows the infundibular cavity in the form of a triangle, 

 but its sides are not quite straight, having a slightly concave form. 

 One corner of the triangle is situated medially and rostrally, the 

 other two corners laterally and caudally. From these latter two 

 corners the canals mentioned run in a lateral direction and curve 

 slightly rostrally. They nearly reach the surface of the brain and 

 are as a rule only separated from the cranial cavity through a thin 

 lamina of nervous matter ; in one case I could not find any nervous 

 lamina at all, but found the canal open into the cranial cavity, only 

 separated by a very thin membrane. 



The »Canalis connectens« runs in nearly straight angle to the 

 »Canalis ventriculi terti« in rostral direction; it is generally some- 

 what curved upwards, but not so much as to prevent the whole 

 canal being seen in a single section. All these cavities and canals 

 are lined with endothelium. The arrangement of the ventricular 

 system resembles much more that which is found in embryos of a 

 certain development of Torpedo than anything found in Petromyzon. 

 Looking at a sagittal section of a Torpedo embryo of 50 mm length 



