421 TRANSACTIONS OP SECTION D. 



and Petromyzontes are non-homologous structures, and, if this be the case, it 

 necessarily follows that in one of the two groups the piston cartilages cannot 

 represent the modified primitive Gnathostome mandible (despite the mandibularis- 

 innervated musculature!). Thus, even according to Ayers and Jackson, the 

 piston cartilage of Petromyzon represents the posterior segment of the basal 

 plate of Bdellostoma, and this, according to them, being composed ' only of 

 chondroidal tissue and . . . not homologous with any part of the visceral 

 arches,' it necessarily follows, on their own showing, that the conclusion just 

 stated applies to the Petromyzontes — the Petromyzontes cannot possess the 

 homologue of a mandible. Further, seeing that the piston cartilages of Myxin- 

 oids are not more obviously a modified mandible than the equivalent struc- 

 tures in Petromyzontes, there is thus left no reason for supposing that the 

 piston cartilages are in either case so derived. Since it is impossible to assume 

 that in one group the mandible and in the other accessory cartilages (e.g., 

 labials) respectively produced the piston skeleton, it must be concluded that 

 jaws were absent in the ancestors of both groups. (4) The assumption of a 

 gnathostomatous Marsipobranch ancestor ignores the primitive membranous 

 condition of the cranium and the feeble development of the parachordals and 

 trabecule, parts which must be strongly built in order to bear jaws, and there is 

 no more reason to suppose that these parts have secondarily degenerated in the 

 Marsipobranchs than it can be supposed that the low degree of cephalisation 

 (i.e., absence of occiput, a feature also correlated with the absence of jaws) is 

 secondary. (5) In view of the undoubted validity of Balfour's well-known dic- 

 tum that ' if the primitive Cyclostomes had not true branchial bars, they could 

 not have had jaws, because jaws are essentially developed from the mandibular 

 branchial arch,' the question as to whether or not the branchial skeleton of 

 Marsipobranchs is homologous with the branchial skeleton of Gnathostomes 

 becomes of importance. Balfour's contention that the Marsipobranchs do not 

 possess and never have possessed a branchial skeleton homologous with that of 

 Gnathostomes is supported by the facts that the branchial basket is developed 

 external to the ventral aorta and the gill-vessels instead of internal as in Gnatho- 

 stomes, that it is developed long before the subocular arch (p.p.q. bar on the 

 hypothesis), and apparently quite independently of it (c/. Gnathostomes in which 

 the jaw and hyoid arches are the first members of the visceral series to appear, 

 even in the degenerate Sturgeon), that it is only developed to any extent in 

 the Petromyzontes, in which the gill-apertures serve for ingress as well as 

 egress of water, and not in Bdellostoma, in which the water enters the gill 

 pouches via the hypophysial duct, and that a series of parietal myotomes 

 extend underneath the entire series of gill-apertures to the mouth in both 

 Myxinoids and Petromyzontes (the presence of these myotomes implying that 

 vertically elongated gill-clefts, and therefore elongated segmented branchial bars, 

 were never present in Marsipobranchs). Viewing the system of visceral arches 

 as a whole, it is incredible, if Marsipobranchs have originated from a gnatho- 

 stomatous stock, that the first two visceral arches should exhibit the differences 

 in development (in time and form) and relationships to nerves and muscles 

 that the subocular arch, piston cartilages, styloid cartilage, &c, do when com- 

 pared with the jaw and hyoid arches of Gnathostomes. If ancestral Marsipo- 

 branchs possessed jaws, they were not the jaws of existing Gnathostomes. The 

 only alternative to the branchial arch theory of the origin of jaws is the cirrho- 

 stomial theory (Pollard), but to adopt this is to adopt Balfour's contention. 



6. Discussion on the Origin of Mammals. 

 Opened by Professor G. Elliot Smith, M.A., M.D., F.R.S. 



The importance of recent acquisitions to our knowledge of comparative 

 anatomy, embryology, and palajontology is the justification for reopening the 

 discussion of this much-discussed problem. 



My excuse for venturing upon the task of dealing with a subject that calls 

 for an intimate acquaintance with a wide range of biological investigation, to 

 which I cannot pretend to lay claim, is the fact that in all previous discussions 

 the consideration of the influence exerted by the evolution of the brain in making 



