636 MR. W. G. EIDEWOOD ON THE [NoV. 6, 



Huxley (6) was therefore unduly confident in assuming that this 

 tubercle was the hyomandibular cartilage which he had discovered, 

 and in writing (p. 34) that "this is neither a process of the 

 suspensorium, nor does it articulate with, nor take the principal 

 share in, suspending Hy, which is Dr. Giinther's ' hyoid arch.' " 



The hyomandibular is still present in this specimen, and lies 

 dorsaUy to the " tubercle," to which a portion of the ligament stiU 

 remains attached. 



The lower end of the hyomandibular cartilage, which Huxley calls 

 the symplectic, is not embedded in the hyosuspensorial ligament, 

 but, as a careful dissection will show, is articulated to the dorsal 

 surface of the suspensorial tubercle, whereas the hyosuspensorial 

 ligament is attached to its posterior surface at a slightly lower 

 level. 



Gadow's replacement of Huxley's accurate term " hyosuspen- 

 sorial ligament" by "ligaraentum hyomandibulo-hyoideum " is 

 thus singularly unfortunate. 



The hyomandibular is not connected directly with the ceratohyal, 

 but indirectly by means of a process of the palato-quadrate 

 cartilage, so that, by a slight stretch of the imagination, the skull 

 of Ceratodus might be regarded as partially hyostylic ; the bulk of 

 the evidence, however, is distinctly opposed to this view. 



Briihl (2) (Taf. Ixiii. Fig. 6), in a figure said to be copied from 

 Huxley, shows a ligament running from the ceratohyal to the 

 mandible and to the opercular, but not to the suspensorium nor to 

 the hyomandibular, and in his original figures he is hardly more 

 fortunate. 



Assuming the determination of the ceratohyal and hyomandibular 

 correct, it is worthy of note that the region of union of these 

 elements, in the Sharks just behind the mandibular articulation, is 

 here, in Ceratodus, raised considerably above the level of the 

 quadrate condyle of the upper jaw (see fig. 2). 



The shape of the hyomandibular is by no means constant, but is 

 subject to great individual variation, and this even on the right 

 and left sides of the same skull. 



By far the commonest form is that shown in fig. 3, a (p. 637). The 

 cartilage is rhombic in shape and applied by its anterior edge to the 

 cranial cartilage, while a ventral process, separated from the skull 

 by a space for the passage of the hyoidean branch of the seventh 

 nerve, is attached to the upper surface of the suspensorial tubercle. 



In one case examined (fig. 3, b) there is in addition a nodule of 

 cartilage in the hyosuspensorial ligament which may possibly 

 have the value of an interhyal. A similar nodule is mentioned by 

 Pollard (10), but not figured. 



The two sketches, fig. 3, c and d, show what variation may 

 occur in the relative lengths of the vertical and horizontal diameters 

 of the cartilage, and also that when the longer diameter is hori- 

 zontal the extent of the attachment to the cranium is reduced to a 

 minimum, 



