246 S. W. WILLISTON. 



striction and union with the short basipterygoid processes of the 

 basisphenoid, the posterior plates of the pterygoids diverge to the 

 inner side of the quadrates, opposite the inner angle of the artic- 

 ular bone of the mandible. These plates have a thin, straight, 

 horizontal lower margin, whence they slope inwardly and up- 

 wardly into a somewhat concave surface, narrow and nearly ver- 

 tical in front, widened behind, leaving a small oval space anteri- 

 orly between them and the basisphenoid. The basisphenoid is 

 narrow in front, gently widened behind, shallowly concave in the 

 middle and limited on either side by a rather rugose ridge. From 

 near the posterior, somewhat divergent, ends of these ridges, a 

 slender bone runs outward and backward by the side of the inner 

 margin of the pterygoid plate. It is in all probability the stapes. 

 Upon the whole, the structure of the palate and occipital region 

 is quite like that of L. hamatns, which will be fully described and 

 illustrated in a future paper from a remarkably perfect specimen 

 in the University collections. Lying under the palate were two 

 slender rods. One of these has been necessarily destroyed in 

 getting to the palatal surface ; the other still remains in the matrix. 

 They nearly meet in the middle anteriorly, just back of the inter- 

 pterygoid vacuity, diverging posteriorly nearly parallel with the 

 margins of the pterygoid plates to near the quadrate. Back of 

 these and articulated with them are a pair of more slender bones 

 bent inwardly near their middle, and terminating acutely behind, 

 nearly parallel with the front margins of the clavicles. 



Vertebra. The vertebrae are united in a continuous series 

 from the skull to near the tip of the tail, forming a U-shaped 

 curve. The pectoral girdle still conceals the anterior five or six 

 of these and the posterior seven of the presacral series were lost 

 in the missing fragment of the nodule. However, the top of the 

 nodule over these missing seven vertebras has preserved in part 

 impressions of them, and, inasmuch as the vertebras themselves, 

 where exposed, are all of precisely the same length, seven and a 

 half or eight millimeters, the number of presacral vertebras in the 

 entire series is determinable with but slight chance of error. This 

 number was either twenty-three or twenty-four. Broili has deter- 

 mined the number of presacral vertebrae in Labidosaurus hatnatus 

 as either twenty-four or twenty-five ; twenty-three or possibly 



