974 



PISCES. 



every degree of formation from a simple furrow 

 to a perfectly enclosed tube. 



The broad scales or plates which form the 

 armour of the Syngnathidse are traversed by 

 similar canals enclosed on all sides, which, al- 

 though separate at the margins, anastomose 

 freely towards the middle of the scale. 



The longitudinal lines in all their forms con- 

 stitute a series of hollow tubes that must be 

 regarded as true canals. These canals traverse 

 the scale in a longitudinal direction converging 

 towards a focus, which, as will be shown here- 

 after, is a centre of nutrition, so that most pro- 

 bably these tubes perform the functions of 

 nutritive vessels. 



2. The cellular Unes, supposed by preceding 

 writers to be merely lines of growth analo- 

 gous to those observable upon the exterior 

 of a bivalve shell, attentive examination shews 

 to owe their arrangement to a very different 

 cause, originating in the developement of pri- 

 mitive cells, which are developed in the super- 

 ficial stratum of the scale and gradually assume 

 an elongated form, become filled with corneous 

 matter, and ultimately arrange themselves in 

 concentric lines of greater or less breadth, 

 which only indicate by their uneven edges 

 their original nature. 



3. The corpuscles seen in scales are precisely 

 similar in their appearance to those met with 

 in bone and cartilage, and are obviously of the 

 same nature. They are distributed in the basis 

 membrane, which seems to be an amorphous 

 tissue resembling that in which the corpuscles 

 of bone are deposited, and which forms the 

 superficial stratum of the scale. 



4. Theorems layer. On scraping off with 

 a penknife the external surface of the scale, the 

 cellular lines and the corpuscles with their 

 basal membrane are removed, and the deep- 

 seated stratum of the scale becomes visible, 

 which is then seen to consist of fibrous lamellae 

 composed of fibres that cross each other at 

 regular angles, giving to this tissue the appear- 

 ance of fibro-cartilage. This layer is thickest 

 near the focus of the scale, and become gradu- 

 ally thinner towards the edges. 



5. The focus is the space towards which the 

 longitudinal lines converge, but is not always 

 situated in the centre of the scale ; it is occu- 

 pied by very large pale corpuscles and by 

 interrupted circular lines; such at least is its 

 appearance in the scales of Acanthopterygenous 

 Fishes, but in the Malacopterygii, and more 

 especially in those species which have mem- 

 branous scales, it presents nothing but a smooth 

 circumscribed surface without corpuscles or in- 

 terrupted laminae, and is then generally sur- 

 rounded by the concentric cellular lines. 



6. The teeth of the scales. The growth of 

 the spines and other appendages seen upon the 

 outer surface and posterior margin of many 

 forms of scales, more especially in those named 

 ctenoid by Agassiz, is a subject of very consi- 

 derable interest, and to the old physiologists, 

 who believed all scales formed by mere exuda- 

 tion, must have been quite unintelligible. The 

 production of these spines is in fact, according 

 to the researches of Dr. Mandl, in every respect 



the growth of teeth, each being 

 a distinct capsule and developed 



similar to 



enclosed in 



in the following manner, as exemplified by 



the growth of a scale of Corvina 7iigra, one of 



the Scienidae. 



The posterior margin of one of these scales is 



occupied by conical appendages, represented in 



a highly magnified condition in Jig. 508. Each 

 f tne long processes 

 here depicted is seen to be 

 enclosed in an envelope, 

 from which, however, it is 

 entirely separate, as is pro- 

 ved by the fact that when 

 the capsule is ruptured the 

 enclosed spine can be re- 

 moved from it with the ut- 

 most facility. Examined in 

 detail, every one of these 

 spines exhibits an organi- 

 zation and mode of growth 

 precisely similar to that of 

 a tooth, being formed in its 

 capsule exactly in a similar 

 manner. The germ begins 

 gradually todevelope itself; 

 it acquires roots, and be- 

 comes distinctly composed 

 of different layers, so that 

 these spines may with pro- 

 priety be called the teeth of 

 the scales, in allusion to 

 the mode of their develope- 

 ment. The marginal teeth 



A small portion of the are most developed, and 

 scale of Cormna ,- h nearest h focus j 

 ara, liiqhlu maqni- , . ,, 



Jfted, shewing the so - Thus m the fi g ure the 

 progressive develop- two superior teeth are seen 

 ment of the teeth enclosed in their capsules, 

 upon its surface. t }, e i r Dases thick and well 

 (After Mandl.) formed> and the whole sur . 



face of the tooth smooth and continuous. In the 

 two next teeth below, the developement is much 

 less advanced; the extremities are truncated, the 

 external layer of the tooth does not entirely cover 

 it, but the roots are visible. Still lower down 

 the teeth of the scale become more and more 

 imperfect, until the lowest are scarcely at all 

 developed, and are barely distinguishable among 

 the surrounding corpuscles. 



In other families with denticulated scales 

 the growth of these appendages is precisely 

 similar, as in the Gobioides, Percoides, Pleu- 

 ronectidse,&c. 



From the above observations it becomes evi- 

 dent that the scales of Fishes can no longer be 

 regarded as mere productions of secretion from 

 the skin, but must be considered as possessing 

 an inherent power of nutrition and a true 

 growth. The denticles which exist upon many 

 of these scales offer by their successive deve- 

 lopement a striking proof of this important 

 fact; while the canals whereby they are tra- 

 versed, and the corpuscles belonging to their 

 structure, plainly intimate that their mode of 

 developement is similar to what exists in the 

 teeth and in the osseous system. 



The chemical composition of the scales of 

 Fishes, moreover, very nearly approximates 



