638 MB, W. G. BIDEWOOD ON THE [XoV. 6, 



hyoidean nerve, instead of lying between the hyoraandibular and 

 the cranium, penetrates the cranium itself, and the hyomandibular 

 is separated by a wide gap from the tubercle to which the hyo- 

 suspensorial ligament is attached. 



It is not impossible that the bar of cartilage behind the nerve- 

 aperture is the symplectic process, which has fused above with the 

 cranium and below with the tubercle (which latter in the specimen 

 represented in fig. 3, e, is practically obsolete) ; but it is more 

 reasonable to suppose that the nerve has been enclosed by the 

 posterior union of the edges of the notch in the cranial carttfage 

 in which it originally ran, parallel cases of the enclosure of nerves 

 and blood-vessels being not uncommon. 



Little doubt can be entertained of the fact that the whole 

 branchial and hyoid system has become reduced in Ceratodus in 

 relation with the partial adoption of pulmonary respiration ; and the 

 hyomandibular thus furnishes another and a very interesting 

 example of individual variation in vestigial structures. 



This reduction of the branchial skeleton is even more pronounced 

 in Proto/derus, where the hyomandibular is absent and where only 

 the second and third branchial arches have epibranchial elements, 

 the other four consisting of slender cerarobranchials only. 



The hyoid gill-filaments in Ceratodus are set on the ventro- 

 internal edge of the ceratohyal, and the series extends backwards 

 and upwards, internal to the hyosuspensorial ligament, but stop- 

 short for about a quarter of an inch in the region of the hyoman- 

 dibular and is continued again in a line running back, parallel to 

 the long axis of the body, to meet the upper end of the series of 

 gill-filaments of the first branchial arch. Thus not only is the 

 continuity of the gill broken in the region of the hyomandibular, 

 but the direction of the series of filaments becomes changed. 



The gill-rakers, on the other hand, form a continuous series run- 

 ning along the upper and inner edge of the ceratohyal, curving 

 forwards and upwards in front of the hyomandibular, and termi- 

 nating at the upper extremity of the first branchial cleft. 



While the series of hyoid gill-rakers and gill-filaments are close 

 together along the inner surface of the ceratohyal, they become 

 widely divergent above, the upper end of the first branchial arch 

 being situated some little distance behind the pharyngeal opening 

 of the cleft. 



The upper gill-filaments are not supported by any part of the 

 hyoid skeleton, but simply project from the mucous membrane and 

 readily come away when the latter is stripped off, and the lower 

 filaments, although supported by the ceratohyal, are but very feebly 

 attached to it and leave no marks on removal, the gill-filaments of 

 Ceratodus not being supported by gill-rays. 



As before mentioned, the hyoid arch of Ceratodus is less reduced 

 than in the other living Dipnoi. 



In Protopterus there is a pedicle for the attachment of the 

 hyosuspensorial ligament, situated much as in Ceratodus, and to 

 the inner side of this short ligament lies a much longer and whiter 



