THE NERVOUS SYSTEM 125 



are scarce, and an outer alveolar layer. The blood-vessels lie 

 between the two. 



The space between the dura mater and the primary vascular coat 

 is called the, primary subdural space. It is not a homogeneous cavity, 

 but is filled with a loose vacuolated tissue which Miss O'Neill likens 

 to embryonal 'Schleimgewebe', so that the subdural space in the 

 Salamander is represented by a number of small spaces rather than 

 by a continuous cavity. 



Miss O'Neill considers this condition primitive, representing a 

 phase earlier than the formation of a true 'arachnoid' membrane. It 

 should be noted here that, owing to the shrinkage of the neural 

 tissue in the processes of fixing and hardening, whether for section 

 cutting or dissection, the primary subdural space is always seen more 

 or less badly distorted. It is therefore difficult to determine its exact 

 relations with accuracy. 



Modifications of the Envelopes surrounding the Brain. The 

 splitting of the dura mater and the formation of an interdural space 

 in certain regions have already been mentioned. This occurs only in 

 the cranium, and does not extend into the vertebral canal. There are 

 four such splits, and they are formed by: (i) the saccus endolympha- 

 ticus; (ii) the saccus perilymphaticus; (iii) the pituitary body, or 

 hypophysis cerebri; and (iv) the paraphysis. 



(i) Spatium interdurale endolymphaticum (O'Neill), as the first is 

 called, is the largest. The saccus endolymphatkus is easily seen, on 

 removing the roof of the cranium, lying at the sides of the midbrain, 

 owing to the fact that it is filled with a white chalky material. It is 

 connected with the ear by means of the ductus endolymphatkus^ which 

 passes through the cranial wall into the auditory capsule. Both Hasse 

 (1873) and Rex (1893) state that it is subdural in position, that is, 

 they imagine that the ductus endolymphaticus penetrates the dura 

 mater and swells out into the saccus endolyphaticus between this 

 envelope and the inner vascular coat. This would put Salamandra — 

 and Triton — in a peculiar position, since, in the Frog, both Coggi and 

 Gaupp had shown the endolymph sac to be i«/«?rdural. Miss O'Neill 

 has demonstrated that this anomaly does not exist, and that in Sala- 

 mandra the endolymph sac splits the dura into two layers, and is 

 therefore interdural as in the Frog.^ The dorsal and lateral walls of 

 the endolymph sac are richly vascular, while the mesial wall — the 

 'pars media' of Hasse — lies very close to the brain and to the plexus 

 covering the fourth ventricle, so that in the region of this plexus the 

 neural layer of the dura and the primary vascular coat are brought 



' For other views on the position of the saccus endolymphaticus see table on p. 128. 



