ago BULLETIN OF THE BUREAU OF FISHERIES. 



and a few other observations not closely allied to the main subject. These investigations 

 have been carried on with Cynoscion regalis and Orihopristis chrysopl-eriis, the scales of 

 which have not been hitherto investigated, with the hope that the results might broaden 

 the knowledge of scales, either by corroborating, modifying, or contradicting the extant 

 theories. 



There are also some observations of the radii with a discussion in which a conclu- 

 sion is reached that is quite at variance with all previous theories of their origin. If 

 this conclusion is sufficiently borne out by facts, it will either negative or seriously 

 modify systems of classification employing the radii as characters. 



A review of literature is necessary in order to bring out the investigations in the 

 light of what has already been done. 



REVIEW OF LITERATURE. 



F.'^STENING TO INTEGUMENT. 



Peters (1841) was the first to devote his attention to scales as a part of aninteg- 

 umental organization. He gives the following analysis of the skin: (i) Epidermis 

 composed of squamous cells; (2) layer of pigmented cells; (3) skin proper, a layer com- 

 posed of fibrous connective tissue and containing fatty globules; (4) an exceedingly 

 thin membrane on the exterior surface of the scale, but distinct from the skin to which 

 it is intimately fastened. In this are found the circuli and radii. He maintains that 

 scales are not found on the epidermis, but in the skin itself. 



Baudelot (1873) described scales as contained in sacs and more or less visible to 

 the exterior, but in some cases (eels, etc.) covered entirely in the skin. The epidermis 

 sometimes extends so far over the posterior field as to be pierced by the teeth in cases 

 of ctenoid scales. The degree of firmness of anchorage to the scale pocket varies from 

 species to species. In imbricated scales, the free portion has intimate connection with 

 the skin. In saying that they are contained in dermal sacs, he implies an agreement 

 with Peters as to their dermal origin. 



Vogt (1842) advances an interesting theory as to the nature of the scale pocket. 

 He regards it merely as a fold in the epidermal membrane. By this he implies that scales 

 have their origin in the epidermis. 



Leydig (1851) says: "The scales of most of our fresh-water fishes appear partly 

 as ossifications of flattened skin continuations which are generally termed 'scale 

 pockets.'" This is close to Vogt's theory, but he confuses "skin" with epidermal 

 folds. 



In considering the work of Klaatsch (1894), done on trout, Esox, and several 

 c)T)rinoids for the 3'ounger stages, I can do no better than quote what he has to say 

 about the fastening to the integument: 



Under the epidermis, which contains a large number of mucous cells, the dermis is seen to be raised 

 in a series of projections, each of which corresponds with the posterior free end of the scale. Each scale 

 lies in an oblique direction from behind forward and becomes inclosed in a compartment of the dermis, 

 the so-called "scale pocket." In this scale pocket one distinguishes an outer and an inner wall. The 

 outer wall consists, in its posterior part, of loose connective tissue containing numerous chromatophores; 

 in the anterior part, the outer wall is composed of tense connective tissue which is similar to the inner 

 wall of the adjoining anterior scale pocket. The fibrous projections of this connective tissue of the outer 



