220 Bulletin American Museum of Natural History. [Vol. XXVII, 



fundamental similarities in Marmosa and Didelphis. These two types further 

 agree in the dental formula, general form and arrangement of the teeth and 

 in the inflected angle of the mandible. The orbitosphenoid is very small 

 and lies above and in front of the sphenorbital fissures which are confluent 

 in the middle line; the vertical plate of the palatine is large, the alisplienoid 

 is large and extends on to the glenoid surface of the squamosal. Internally 

 to the post- glenoid process the alisplienoid sends back an inflated expansion 

 the " alisphenoid bulla.'" This is seen in a primitive stage in the genera 

 under consideration, where it merely embraces the anterior part of the 

 membranous tympanum. The malar in both genera is very large and has 

 a prominent postorbital apophysis which may possibly be a remnant of the 

 postorbital bar in the Cynodonts (p. 120). The malar extends back to the 

 o-lenoid fossa and has an articular surface which limits the lateral vibration 

 of the mandibular condyle (p. 1 19). The zygomatic process of the squamosal 

 is relativelv short and is received anteriorly between the long inferior, and 

 short superior, fork of the malar. The scjuamosal sends back a ridge which 

 runs from the dorsal edge of the zygoma, above the external auditory arch 

 to the lambdoidal crest. This ridge is analogous to a similar one in Cyno- 

 gnathus (p. 121). In Marmosa it is but feebly indicated. The orbital portion 

 of the lachrijmals is large but the facial portion in Marmosa is very limited; 

 the lachrymal foramen is marginal. The pterygoids are reduced to thin 

 scales of bone and in this respect Marmosa and Didelphis both appear to 

 have become more specialized than Thylaciinus (Fig. 1, A). The palatal 

 fenestra' are prominent in both genera. Parker (1886, p. 270) found that in 

 the develoj^ing skull the palatal plates of the maxillary and palatine bones 

 in Marsui)ials generally were at first not fenestrated but became so by the 

 o-radual absorption of their substance in certain areas. This conspires with 

 other evidence to indicate that the fenestration of the palate in Marsupials 

 and Insectivores is a secondary character. The posterior part of the palate 

 in 1)oth Marmosa and Didelphis terminates in a prominent transverse 

 palatal ridge, which is pierced at its opposite external corners by a prominent 

 foramen, which is possibly for the posterior palatine nerve (a branch of V2). 

 The internal view of the skull of Didelphis also reveals many primitive 

 mammalian characters. The cerebral chamber is longer than high, the 

 olfactory fossa is large and long; the long axis of the brain case makes only 

 a very gentle angle with the nasal cavity. Beginning at the posterior end 

 of the skull we observe the following structures : The paired condylar fora- 

 men, piercing the exoccipital; the foramen lacerum posterivs (f. jugulare), 

 h'ing between the basioccipital and the periotic; the internal auditory meatus 

 (for the facial and auditory nerves) piercing the petrous; and above it the 

 circular floccular fossa. Running through the occipital wall of the supra- 



