246 Bulletin American Museum of Natural History, [^^ol. XXVII, 



Centetes it is separated from the basioccipital by the backward extension of 

 the tympanic flange of the basisphenoid. On its inferior surface it also bears 

 a depression which is apparently caused in ontogeny by the pressure of the 

 tympanic chamber, as shown by comparison with Microgale (Fig. 18, B). 

 In all these details the petrosal of Soicnodon resembles that of Ictops. The 

 petrosal of SoJenodon paradoxus sends out a small postero-external wing 

 (Fig. 18, A^, pr. ty. pet.), which in the adult articulates with the postero-inter- 

 nal border of the tympanic. A similar process was observed by Brandt in 

 S. cuhaniis (van Kampen, 1905, p. 425) and is also represented in Micro- 

 gale (Fig. 13, B). There seems strong evidence for homologizing this tym- 

 panic wing of the petrosal in Centetoids with the entotympanic of INIarsupials 

 and Menotyphla (van Kampen, 1905, p. 452). 



The mastoid portion of the periotic is relatively large in the young Soleno- 

 do7i, and is broadly continuous with the petrous. It faces obliquely outward 

 and backward and is well seen in the side view as in Microgale (Fig. 18), 

 and Potainogale, whereas in Ictops and Centetes it faces backward rather 

 than outward, and is seen best in the occipital \ iew. On its inferior surface 

 the mastoid has a deep circular ])it or notch (Fig. 18, A^,f. st. m.), which may 

 have served for the attachment of the stylomastoid bone. A similar pit 

 occurs in Centetes and Erinaceus. The ring shaped tympanic is relatively 

 large and broad. Anteriorly it rests against the postglenoid process of the 

 squamosal (contrast Centetes). It is o]:)lique rather than horizontal or vertical 

 in position. 



Nerve foramina. The foramina are best demonstrated from an internal 

 view of the skull in sagittal section. In the olfactory foramina, piercing the 

 cribriform j^late nothing remarkable was noted. The ethmoid foramen, for 

 the recurrent, ethmoid branch of the ophthalmic division (VJ of the trige- 

 minus, as in Didelphis and Placentals, is located just behind and externally 

 to the inferior part of the cribriform plate. This foramen issues externally as 

 in Fig. 18, A^ (/. etJi.). Behind and much below and internal to the ethmoid 

 foramen and also internal to the large sphenorbital fissure is the minute 

 optic foramen, piercing the orbitosphenoid as in typical Placentals and 

 agreeing in its small size with that of Centetes, and Potamogale. The fora- 

 men laccrum auterius (f. sphenorbitale, nerves III, I^^, Vj, VI) is very large 

 and circular. Some distance behind it, and perforating the base of the 

 alisphenoid, is a small foramen which at first sight looks like the foramen 

 rotundum. But this foramen faces backward, and therefore cannot trans- 

 mit the superior maxillary branch (V,) of the trigeminus. Moreover it 

 issues externally in the posterior opening of the alisphenoid canal and nuist 

 therefore have transmitted a branch of the external carotid artery. The 

 inference is that the for aiyien rotundum in SoJenodon, as in Centetes and other 



