HATCHER : OLIGOCENE CANIDiE 



73 



that portion of the petrosal enclosing the cavity of the internal ear, as shown on 

 the left side in Fig. 2, we shall find a very close similarity between that structure 

 and that which has been described by Leidy and Scott as the auditory bulla. The 

 minute external opening described by Scott is no doubt the fenestra rotunda of the 

 petrosal and not the external opening to the auditory bulla. The reniform fossa 

 noted by Leidy is present also on the postero-internal side of the petrosal, though 

 not so pronounced in recent species of Canis. Not only do all these and many other 

 characters go to show that this element is in reality the petrosal, but a fragment of 

 the tympanic has been retained in skull No. 492 and is shown in position and over- 

 lying the petrosal in Fig. 3. A careful comparison of Figs. 2 and 3 will make it 



Fig. 2. Inferior view of basicrauial region of skull of Canis familiaris with auditory bulla removed from left side 

 to show structure of inner ear for comparison with Fig. 3; r/., reniform fossa; pr., promotory ; f.r., fenestra 

 rotunda. 



apparent that the bone in question is the petrosal and that through imperfect con- 

 nection with the surrounding elements the auditory bullae have very generally been 

 entirely lost in Daphwnus during the process of fossilization. Considering this in 

 connection with the fact that in several species of recent dogs, even in adults, the 

 auditory bullae are easily detached, this may be considered as a primitive condition 

 among the canidse, while the completely ossified and firmly ankylosed auditory 

 bullae should be regarded as more specialized characters. 



The petrosal therefore is the element which has been mistaken by Leidy and 

 Scott for the auditory bulla. It is proportionately smaller than in the modern 



