THE SHOULDER-GIRDLE IN FISHES. 55 



from the outer or convex side; and fig. 11 is a section through the widest part, illustrating the 

 convexo-concave character of the plate, and the manner in which it is becoming ossified. The 

 general form is crescentic, but the upper and lower ends are dilated the former in a rounded 

 manner, and the latter, which is much less enlarged, is ernarginate. The middle part of the 

 cartilage is greatly expanded ; in front in an evenly rounded manner, whilst the hinder margin is 

 developed into two slightly separated lobes. 



As there is, in my knowledge, no Fish with a fenestrate coracoid, I shall take the " fenestra " 

 as a landmark. To the middle of the " fenestra," at least, and perhaps lower, all is referable to 

 the scapula ; and the osseous plate is essentially scapular, but the fenestra is coraco-scapular, as in 

 the Codfish ; and the scapular bony plate has trespassed on to the coracoid region. But the whole 

 of the glenoid margin (gl.) is scapular, and below that region to some extent ; here the bony sub- 

 stance began, and passed over the inner and outer side of the plate simultaneously. The glenoid 

 region has four irregular notches, and there is a very arrested brachial series of rays articulating 

 with that notched border. The whole brachial plate has the form of a triangle, nearly equilateral, 

 with the apex downwards. From the base, one somewhat bent, drumstick-shaped ray (b. 1), 

 with soft ends, has been perfectly cloven ; but the next ray, which is shorter, is separated from 

 the one below it by a long " notch," which is closed proximally. A feebler and shorter third 

 ray, not severed from the second, comes next ; this is severed from the rather large apical part by 

 an anterior notch, closed by a considerable isthmus of cartilage behind. This band of cartilage 

 is continued nearly all round the lowest and smallest ray (b. 4), which is triangular, and clamps 

 the upper edge of the apical moiety. 



The morphologist could not desire a more instructive stage than this ; the process of cleavage, 

 or dehiscence, takes place, as it were, before his eyes, and in obedience to a law which in one 

 sense is unalterable, and, in another view, may be said to produce all the variety to be seen in 

 the Vertebrate Endo-skeleton. 



Fam. 

 Example. Salmofario, Linn. 



In Plate II, figs. 7 and 8, the Shoulder-girdle of a half-grown Trout are shown, magnified 

 two diameters ; fig. 7 shows the outside of the left moiety, and fig. 8 the inside. The splint bones 

 are simple, fibrous plates; tender, and very inornate, as compared to what we see in the 

 Acanthopteri. The post-temporal (p. t.) is strongly forked ; the supra-clavicle (s. cl.) long and 

 spatulate ; the post-clavicles flat, thin, and, contrary to the rule, the lowest is much the largest. 

 The clavicle (cl.) is strongly bent forwards on itself at the upper third ; its post-branchial plate 

 forms almost the whole of the bone, the outer part of which is parallel with the sides of the Fish, 

 whilst the inner part, slightly concave, looks as much forwards as inwards. The posterior plate, 

 therefore, must be looked for on the inner side (fig. 8) ; it is a very long balk or ridge of -bone, 

 and gives attachment to the anterior margin of the scapula. This latter bone (sc.) is imperfect 

 in front, there being a large headland of soft cartilage; it has in it a smallish " fenestra" (sc. f.) ; 

 the whole region is subquadrate. Between the scapula and the coracoid proper there is a 

 considerable tract of cartilage (fig. 7, sc. cr.) ; the coracoid (cr.) is thick, trihedral above, 

 elegantly arcuate, with the convexity backwards, and it has a long sub-brachial process ; both the 



