THE ELASMOBRANCH FISHES 67 



seventh eeratobranchial practically at right angles to its long axis. These rays 

 extend posteriorly between the plate x and the median piece as clear pieces of 

 hyalin cartilage. Whetlier they represent rudimentary branchial rays on the 

 seventh arch like those described by Gegenbaur (1872, pi. 12, fig. 5) on the 

 anterior margin of the fifth arch for Scjjllium, or have to do with the riidi- 

 mentar}^ arch following, is not certain. 



On the middle piece a similar arrangement is found. Here there are three 

 pieces which are successively longer toward the middle line. They are essen- 

 tially identical in appearance with the rays above described on the seventh 

 eeratobranchial segment, but they are attached along their whole dorsal 

 length as flattened lamellae. Terminally the median two look very much like 

 the rays above described and they are much like them also in that they are of 

 clear hyalin cartilage. Farther toward the median line on the middle piece 

 there is clear evidence of still another similar group, except that it is no4eepa- 

 rated into rays or lamellae. This group is also of clear hyalin cartilage aiff er- 

 ing distinctly from the median piece, which, in the specimen, is of a dark 

 color. I have interpreted this group as a remnant of a ninth arch (ar.^), al- 

 though I am not certain what part it represents. 



The condition found in this specimen is suggestive as to the method of for- 

 mation of the enlarged median piece so characteristic of the Elasmobranchs. It 

 would appear that in this region the rudimentary arches are forced more and 

 more to take a longitudinal direction nearer the middle line, and that the 

 median piece represents in its most posterior part the fusion of these arch^^ 

 from side to side. 



Extending from all the visceral arches, except the first (mandibular) and 

 the last, is a series of cartilaginous branchial rays for the support of gills. 

 These supporting rays are confined to the epi- and cerato-segments. The bran- 

 chial rays may be complex and branched, the termini fusing into arches on 

 the liyoid (sharks) , or they may be comparatively simple and straight (rays) . 

 In Torpedo a curious modification of the branchial rays occurs in the form of 

 terminal discs (&.r., fig. 63), each of which is almost circular in shape. These 

 cartilaginous rays, although fewer the more posterior the arch, may be ex- 

 ceedingly numerous anteriorly as in Lamna, or relatively sparse as in Lae- 

 margiis. The central ray, the one between the epi- and the eeratobranchial, 

 may exceed all others in length. This ray may be postulated as the main axis 

 of the fin skeleton according to the gill-arch theory for the origin of paired fins. 



EXTEAVISCEEAL AECHES 



Outside of the deep internal visceral arches are the extravisceral cartilages. 

 These may be divided into the labials^ of the mandibular arch, the extrahyals 

 of the hyoid, and the extrabranchials of the branchial arches. Normally each 



1 The labial segments are often interpreted as representing arches formerly present be- 

 tween the mandibular and hyoid arches. 



