1010.] Hoiv did the Quadrate become free? 139 



sidered as even probable: (1) How did the reduced (juadrate, which is gen- 

 erally spoken of as being "fixed" in the Cynodontia, become sufficiently 

 free from the scjuamosal to enable it to assume the functions of the incus? 

 and (2) how did the reduced articular gain contact with the stapes without 

 interfering with the hearing function? 



(1) The ([uadrate in Galesaurus and Cynognatluis {cf. Gadow, 1901, fig. 4; 

 Broom, 1904, 2, pi. xxxv) appears to be fixed in the squamosal by its dorsal 

 portion {cf. Fig. \, B, Qn.), but even here it seems possible that the low^er 

 end may have been capable of a slight antero-posterior motion. A little 

 atrophy of the posterior wall of the descending flange of the sf[uamosal, 

 which is certainly very thin at this point, would greatly increase the mobility 

 of the lower end of the (luadrate. In Gomphognathus kannemeyeri (Broom, 

 op. cif., pi. xxxv, figs. 2, 3) the cpiadrate is a small bone flattened antero- 

 posteriorly; in the front view it appears to be held in place against the 

 squamosal only by the articular, and that too only on its inferior surface, and 

 there is nothing in Broom's figures to show why these two bones together 

 may not have moved more or less against the squamosal. This quadrate of 

 Gomphognathus parallels the incus of primitive mammals in the following 

 particulars: (a) it is a very small flattened bone attached to the scjuamosal 

 and located at the postero-external corner of the skull, just internally to the 

 backward prolongation of the zygomatic arch {cf. Echidna and Parker's 

 section of the skull in the embryo Tatusia: 1886, pi. iv, fig. 11); {h) it 

 articulates by a convex surface with the articular (the supposed homologue 

 of the malleus) and this circumstance suggests the large convexo-concave 

 articulation between the incus and the malleus in embryo mammals. If 

 the quadrate and articular were further reduced in size it seems possible 

 that they might acquire enough mobility to function as the incus and 

 malleus, which indeed move through very small arcs. 



(2) The problem of the manner in which the reduced quadrate and articu- 

 lar could gain contact with the middle of the primitive stapes-extracolumella 

 without interfering with the hearing function is very obscure, and luitil all 

 the mechanical stages are clearly conceived the whole hypothesis must re- 

 main on trial; but whatever the final answer may be it must take into 

 account the following among other considerations: 



(a) The fact that the tympanic cavity arises as a diverticulum from the 

 Eustachian tube, and enfolds the ossicula from below and internally (p. 126). 



(6) The oblique position of the tympanic membrane in primitive mam- 

 mals, and the fact that the membrana propria appears to be an adjustment 

 for regulating the tympanum (p. 127). 



(c) The supposed homology of the extracolumella with the manubrium 

 mallei (p. 132). 



