THE SHOULDER-GIRDLE IN FISHES. 57 



has several enlargements, both before and behind ; one of these, in front, is a " cycloidal" plate 

 growing off from the body of the bone, and attached by a broad isthmus. The post-clavicles 

 (Plate II, fig. 6, p. cl.) are extremely interesting, for there is an additional piece ; and, whilst 

 they retain their normal relations with the supra-clavicle, they yet have a new relation to the 

 clavicle itself, lying on its outside. 



The first (p. cl. 1) is a long lozenge, lying within the supra-clavicle at the top ; the next 

 (p. cl. 2) is spatulate, and is overlapped in the same way by the one above ; it also overlaps the 

 top of the third (p. cl. 3), which is long, styloid, /-shaped, and reaches the abdominal line. 

 These three post- clavicles answer to a subdivision of the subcutaneous part of the second infero- 

 lateral plate of Callichthys (Plate I, fig. 9, p. cl.) ; they have no representative in Gastcrosteus 

 leiurus (see Woodcut, fig. 5 A) for the second lateral line-bone (1. 1. 2) is small in that Fish, and 

 the rest of that second cincture is aborted by the huge Shoulder-plates. The abnormal position 

 of the post-clavicles in the Herring seems to arise from their having taken the cue of their 

 growth from the supra-clavicle only, thus losing their relation to the clavicle, which, indeed, 

 grows in a very wild way sometimes ; and I believe that variations in overlapping in these splint- 

 bones are merely like variations in the growth of sutural teeth. 



The bend backwards in the clavicle is very low down ; and the scapula is placed below the 

 bend (Plate II, fig. 4, cl. sc.) ; whilst the coracoid is placed face to face against the hollow inside 

 of the lower part of the great splint, and reaches quite as low, namely, to the abdominal line. 



There is but little cartilage left between the coracoid and scapula (fig. 5, sc. cr.), and their 

 line of union is very sinuous ; the front margin of the fenestrate scapula is soft, and so is the 

 space between the styloid prse-coracoid, and the coracoid (fig. 4, p. cr. cr.). The bones them- 

 selves are very strong and ivory-like ; but the coracoid is quite a sieve, all but the margin being 

 pierced throughout. Two of these perforations, in the lower part of the bone, tend to divide it 

 into a prse-, meso-, and post-coracoid ; but the fission is very irregular, and its dehiscing force has 

 been scattered all over the bone. The supero-posterior part of the coracoid is strongly developed 

 into sub-brachial snags. The brachial plate is well developed, its smallest end upwards, as in the 

 Trout ; it is divided both horizontally and vertically ; horizontally into five bars, and vertically 

 into two series, the front rays being much larger than those behind, which are little rounded 

 masses of bone. This subdivision of the arm-bones is what obtains in Polypterus ; it is new to 

 me as occurring in a Teleostean, and I believe, in this respect, new to Ichthyotomy ; at any 

 rate, it is another explanation of the reason why the genus Amia should have such evident 

 Clupeoid relationships. 



