130 THE SALAMANDER 



They are not quite terminal, but arise ventrally a short distance 

 from the anterior end of the prosencephalon. 



(ii) The floor of the thalamencephalon (the.) fits in as a wedge 

 between the hemispheres, which are to be regarded as lateral out- 

 growths from it.^ The optic nerves (n.2) pass from its floor at about 

 the level of the posterior ends of the hemispheres. 



(iii) The injundibulum (inf.) is an outgrowth from the floor of the 

 diencephalon which extends in a postero-ventral direction so as almost 

 to obscure the floor of the mesencephalon. Its posterior end is some- 

 what expanded laterally and has a fairly deep median notch. Trans- 

 verse sections show that the neural tissue is almost entirely confined 

 to the ventral surface, its roof being very thin and membranous. In 

 sagittal section it appears as a hollow wedge. From the posterior 

 angles of the median cleft the pituitary body is suspended. 



(iv) The pituitary body (pit.), or hypophysis cerebri, is roughly 

 a broad oval in surface view, with a median notch anteriorly. The 

 lobes at the sides of the notch are distinguished as the pars tuheralis 

 (Atwell) (pit.'"), while the main mass is the/>^rj anterior. According 

 to Sumi the pars tuheralis arises rapidly at about the time of metamor- 

 phosis. In many Urodeles it becomes completely detached from the 

 pars anterior, but in Salamandra its separation is only partial. Both 

 the pituitary and the infundibulum lie within the basi-cranial fonta- 

 nelle of the primary chondrocranium, and are covered ventrally by 

 the parasphenoid bone. 



(v) The ventral surface of the medulla calls for no special comment 

 save that the roots of the trigeminus and acustico-jacialis are here more 

 distinct, and the A^. abducens (VI) is also visible. It arises on either side, 

 very close to the middle line, by two roots about a millimetre posterior 

 to the acustico-facialis. It is an exceedingly fine nerve and readily 

 broken, and even when intact it is not easy to see. 



In lateral view it should be noted that there is only a very slight 

 departure from the straight unflexed condition of the embryonic 

 brain. 



By direct observation of the sagittal section (Fig. 50) it is not 

 possible to make out more than the main features. If the half of the 

 brain is immersed in a solution of ferric chloride the iron salt will be 

 found to adhere more readily to the grey matter than to the white, 

 and thus render the outlines of the cavities more distinct. The posi- 

 tions of the commissures have been confirmed, and in some cases 

 determined, by the help of transverse sections examined micro- 

 scopically. 



' Cf. Graham Kerr on the brain of Lepidosiren, Q.J.M.S., London, vol. 46, p. 448. 



