TRANSACTIONS OF SECTION D. 627 



Pisces into two series, which would associate the teleosts and ganoids with the 

 cyclostomes, as distinct from the rest, receives support from recent study of the 

 head-kidney by a Japanese, who seeks to show that the organ so called in the 

 Elasmobranchs is of a late-formed type peculiar to itself; '■'" and it is also in agree- 

 ment with one set of conclusions previously deduced from the study of the reproduc- 

 tive organs.-'" 



To deal further with the fishes is impossible in this Address, except to remark 

 that recent discovery in the Gambia that the young of the Teleostean genera 

 Heterotis and Gt/onnarchus bear filamentous external gills, renders significant 

 beyond expectation the alleged presence of these among the loaches, and shows 

 that adaptive organs of this type are valueless as criteria of aflinity.^=' 



In palfeontology, as in recent anatomy, our records of detail have increased 

 beyond precedent, often but to show how deficient in knowledge we are, how 

 contradictory are our theories and facts. 



In dismissing the fishes, I wish to comment upon our accepted terms of 

 orientation. To speak of the median fins as dorsal, caudal, and anal, of the pelvic 

 as ventral, and of the pectoral in its varying degrees of forward translocation as 

 abdominal or thoracic, though a convention of the past, is to-day inaccurate and 

 absurd. I question if the time has not come at which the terms thoracic (pulmo- 

 cardiac) and abdominal are intolerable, as expressing either the subdivisions of the 

 body-cavity or anything else, outside the Mammalia, which alone possess a dia- 

 phragm. Even in the birds, to grant the utmost, the subdivision of the coelom 

 if accurately described, must be into pulmonary, hyper-pulmonary, and cardio- 

 abdominal chambers ; while with the reptiles the modes of subdivision are so com- 

 plex that a special terminology is necessary for each of the several types extant. 



In the fishes, where the pericardium is alone shut off, the retention of the 

 mammaliam terms but hampers progress. This was indeed felt by Dumeril, when 

 in 1865 he attempted a revisionary scheme.^'^ Since, however, one less fantastic 

 than his seems desirable, I would propose that for the future the ' anal ' fin be 

 termed ventral, the ' ventral ' pelvic ; and that for the several positions of the 

 pelvic, that immediately in front of the vent, primitive and embryonic (which is 

 the position for the Elasmobranchs, Sturiones, Lower Siluroids, and all the higher 



desirable, now that it is known that a group of Cretaceous fishes (the Cteno- 

 thrissidae), hitherto regarded as Berycoids, are in reality of clupeoid affinity, 

 despite the fact that at this early geological period they had translocated their pelvic 

 tin into the jugular (' thoracic ') position. '"*' 



The sum of our knowledge acquired during the last twenty-eight years proves to 

 us that, among the bony fishes, the structural combination which would give us a 

 premaxillo-maxillary gape dentigerous throughout, a proctal pelvic fin, a heart 

 with conal valves, would be the lowest and most primitive. Inasmuch as this 

 character of the heart, so far as at present known, exists only among the Clupesoces 

 (pikes and herrings and their immediate allies), these must be regarded as lowly 

 forms ; '<" wherefore it follows that the possession of but a single dorsal fin is not, 

 as might appear, a necessary index of a highly modified state. 



Before I dismiss the vertebrates, a word or two upon a recent result of 

 morphological inquiry which concerns them as a whole. I refer to the develop- 

 ment of the skull. Up to 1878 it was everywhere thought and taught that the 

 cartilaginous skull was a compound of paired elements, known as the trabecule 

 cranii and parachordals, and that the former contributed the cranial wall. 

 Huxley in 1874, from the study of the cranial nerves of fishes, had reiterated the 

 suggestion he made in 1864, when dealing with the skull alone, that the trabeculse 

 might be a pair of prse-oral visceral arches, serial with those which support the 

 mouth and carry the gills. The next step lay with the Sturgeon, in which in 1878 

 it was found that the cranial wall is originally distinct.i°- And later, when the facts 

 were more fully studied in sharks, batrachians, reptiles, and birds, it became 

 evident that the trabeculse, though ultimately associated with the cranial wall, 



S S2 



