NO. 2147. TWO EXTINCT MAMMALS FROM TEXAS— HAY. 119 



lower floor of the bullae is missing, so that the form of this part, if 

 ever present, can not be observed. According to the description of 

 these bulloe in the Brazilian species there is along the median line 

 a space only about 5 mm. wide between them. They are evidently 

 marked off along their inner boundary much more sharply than in 

 the Texan species. In this animal there is between them a broad, 

 longitudinal groove whose sides slope downward and outward grad- 

 ually into the walls of the bullae. The length of each bulla is 50 

 mm.; the width may be taken as 35 mm. The distance from the 

 outer wall of one to that of the other is 90 mm. The median side of 

 each cavity extends inward and upward into the base of the skull 

 until the two are only 15 mm. apart. 



The pterygoid bullae of N. graciUceps have been described by 

 Stock. They are called by him tympanic bullae, but they are not 

 such. Mr. Gerrit Miller has directed my attention to similarly placed 

 and apparently homologous cavities at the base of the skull in vari- 

 ous bats. As shown by Stock the roof of these bullae is formed by 

 the alisphenoids. The side walls and floor in the Brazilian and the 

 Californian species are certainly formed by the pterygoids. In N. 

 graciUceps Stock there is along the inner face of the bulla a slit 

 about 30 mm. long which puts the cavity of the bulla in communi- 

 cation with the pharynx. The bulla of N. texanum appears not to 

 have had a floor. The pterygoids seem to form a wall which sur- 

 rounds the cavity on both sides. On the median side the edge of 

 the wall is partly intact, partly injured. On the outer side the 

 w^all comes down to a sharp thin edge which appears to be little if 

 at all injured. In places the edge is certainly wholly natural. Such 

 being the case the bulla is incomplete and is a cavity opening below 

 by a mouth 30 mm. wide. In N. graciUceps the outer wall has 

 grown downwards and inwards until it has nearly met the inner 

 wall; in N. eserivanense the space between the two walls was ap- 

 parently abolished. In Choloepus hoffmanni there are homologous 

 bullae which open at the anterior end into the mesopterygoid fossa. 

 Similarly placed bullae are found in the great anteater {Myrmeco- 

 phaga juhafa), but their structure is somewhat doubtful.^ 



In the Texas specimen there is a rough and sharp ridge which 

 begins on the midline between the front ends of the pterygoid bul- 

 lae and runs forward as far as the bone is uninjured. A similar struc- 



1 From an examination of skulls of the great anteater in the United States National 

 Museum the writer concludes that the pterygoids and the alisphenoids of each side are so 

 completely coossifled that the line of union can not be determined unless it be in younger 

 individuals than are at hand. The bullte in adult individuals are completely closed. In 

 a not fully grown specimen the impression given is that the bullte remained open longest 

 on the outer side, near the border of the temporal bone. It is believed that the area 

 called alisphenoid in Weber's figure 332 taken from Pouchet (Saugetiere, p. 434) is not 

 such. Certainly the foramen ovale pierces the alisphenoid ; and it is this bone, not the 

 pterygoid, which joins the basisphenoid. 



