407 



additional lateral shelf, the slight convexity of which plays upon the 

 jugal and is almost reached by the investing scale of the squamosal. 

 The next step in this direction is a stage in which the mandible 

 articulates with the squamosal. 



Professor Seeley has permitted me to examine his splendid spe- 

 cimens ofGomphognathus Kannemeyeri, of triassic age. The 

 mandible articulates with the quadrate and with the squamosal, while 

 the quadrate itself is reduced to a very small bone. The articulating 

 process of the mandible lies quite external to the dentary and is prob- 

 ably not the articulare but the strongly developed lateral splint-bone 

 or supraangulare. The large coronoid process is not formed by the 

 coronoid, but by the dental, as in Mammalia. 



At any rate we see demonstrated the gradual transference of the 

 mandibular attachment outwards, onto more lateral elements, namely 

 to the jugal and squamosal, and it is not a wild guess that this new 

 connexion is formed not so much by the articulare than by the in- 

 vesting bones; either by the supraangulare as in Tri onyx, or if 

 this be lost, by the dentale. In this way we understand how the 

 quadrate may be relieved of its suspensorial work and be devoted 

 mainly to the tympanum - 

 bearing function, a func- / ^^ 



tion which it fulfills already /rv^Y' / S<i y-<^ 



in Anura, Crocodiles, Che- \ri ( / /^'XS~ f^^-^^"^ 



lonians , and to a lesser j r\ rL_^ r\ uri ' 



extent in other Sauropsida. r^^^^^^^ v^^J I 



Of course this release of ^-^ ^ 



the quadrate cannot take Fig. 5. Evolution of the Mammalian mandibular joint. 



place in creatures like the 



Theromorpha mentioned above, since their quadrate is too much hemmed 

 in and fixed between the pterygoid and squamosal, but this only shows 

 that we have no business to look upon such Theromorpha as the 

 direct ancestors of the Mammalia. 



It is worth noting that the quadrato -jugal has disappeared. Its 

 place is taken by the elongated jugal, clearly so inGomphognathus 

 and Gyno gnathus, a circumstance not favorable to the comparison 

 of the quadrato-jugal with the tympanicum. The small size and po- 

 sition of the quadrate on the other hand might at first sight seem to 

 lend support to Albrecht's and Baur's view, that this element forms 

 the zygomatic process of the Mammalian squamosal. The quadrato- 

 jugal would be more amenable to such a comparison. 



Lastly it has been pointed out triumphantly by Kingsley and 



