TRANSACTIONS OF SECTION D. 423 



trally to the arch instead of dorsally, and that the anterior palatine branch 

 of the trigeminus runs dorsal to the so-called palatine instead of ventral to it, 

 the so-called palatine being a mere anterior extension of the trabecula (Allis). 

 The so-called hyoid process ('styloid' in Petromyzon) also canno't be so termed, 

 since the hyomandibular nerve passes behind the cartilage (Allis). As to the 

 piston cartilage, since its musculature is innervated by the mandibvlaris nerve 

 and not by the hypoglossal, modern upholders of Gnathostome ancestry have 

 revived tlie opinion that the piston cartilage represents the much-modified 

 and displaced mandible of Gnathostomes, the so-called hyoid representing a 

 quadrate element. The grounds upon which this comparison is made are 

 (1) identity of position, (2) innervation of the piston musculature, and (3) the 

 tacit and quite unwarranted assumption that the piston cartilage, if proved to 

 be homologous with Meckel's cartilage of Gnathostomes, was itself a biting 

 apparatus in ancestral Marsipobranchs. With regard to (1) mere position is no 

 criterion of homology, yet, after all, similarity of position is the only basis for 

 the great majority of the assertions of Ayers and Jackson, and Stockard's sup- 

 posed demonstration of the homology solely amounts to proving that the piston 

 cartilage arises in the position in which it is found in the adult ( ! ), no evidence 

 being adduced to snow that the piston cartilage is ever paired (an interior fork- 

 ing cannot be regarded as evidence), that it ever has any connection with the 

 quadrate element, or that it even at any stage borders the sides of the mouth as 

 mandibular elements should. Respecting (2), it is obvious that this evidence, 

 regarded by the authors named as conclusive, is in reality only an additional 

 illustration of the truism that muscles formed from the same region of the gut 

 wall are in all Craniates innervated by the same nerves (strictly speaking, in 

 this particular case, the mandibular nerve supplying these muscles is not exactly 

 homologous with the Gnathostome mandihuJaris — P. Fiirbringer, Johnston). In 

 what manner this evidence can be held to show that the piston cartilage was 

 once a mandible it is difficult to conceive. (3) Requires no comment. The evi- 

 dence adduced in favour of the Gnathostome ancestry of Marsipobranchs being 

 thus inconclusive, and all authorities being in agreement concerning the exceed- 

 ingly primitive nature of the group, it follows that if there exist facts tending 

 to show that jaws could never have existed, great weight must be attached to 

 this evidence. (1) The development of both the piston musculature (Bujor, 

 Dean) and cartilage (Stockard) in the mid-ventral line is quite inconsistent with 

 the view that these structures were formerly paired laterally-placed mandibular 

 muscles and rami. (2) In the Myxinoids there is no hypoglossal musculature 

 or nerve (Worthington, Cole), the myotomes extending laterally in an unbroken 

 series to the extreme end of the head. This persistence of the primitive con- 

 dition (shown to be primitive by the innervation) proves that a jaw apparatus 

 could never have been developed, since in all Gnathostomes the presence of the 

 laterally developed jaws and the large muscles which work them render it 

 impossible for myotomes to persist in this region. On the other hand, a rasping 

 dentigerous piston worked by a mid-ventral musculature does not interfere with 

 lateral myotomes. In the Petromyzontes the development of the gill-clefts has 

 (unlike those of Bdellostoma) interfered with the ventral growth of the myotomes 

 in the branchial region, the result being the development and growing forwards 

 under the row of gill-apertures of a sub-branchial myotome musculature inner- 

 vated by a hypoglossal nerve and homologous with the hypoglossal musculature 

 of Gnathostomes (Neal, Johnston). In Petromyzontes, however, this sub- 

 branchial musculature extends right up to the buccal funnel, retains its parieto- 

 lateral position, and has no connection with a tongue apparatus, being alone 

 concerned with the flexion of the body anteriorly in swimming. These facts I 

 think also prove that in Petromyzontes a jaw apparatus can never have been 

 present, since had this been the case these lateral sub-branchial myotomes would 

 have become much reduced, restricted to the mid-ventral line, and probably 

 specialised to form a true mid-ventral tongue musculature as they have in all 

 Gnathostomes. To assume that they have secondarily reacquired their primitive 

 appearance is unwarranted, and to assume that they never were specialised as 

 hypoglossal musculature is, to judge from existing Gnathostomes, to assume that 

 jaws were never present. Moreover, the presence of a true tongue musculature 

 would have prohibited the evolution of a piston musculature. (3) According to 

 the views held by most leading authorities, the piston cartilages of the Myxinoids 



