﻿100 62 



tissue with some lieeled scales, whicli are considerably larj^er anil esiiecially longer than the 

 rest of the dermal covering"; adding in a foot-note that he has not devoted any further 

 attention to these scales. 



Jordan and Evkrmann (19, p. 758): "Some hony strips on the side of the hack and on the 

 margin of the thorax and ahdomen, the former sometimes confluent into a shield". Quite 

 the same is said by Jordan and Starks (20, p. 68 and 69). 



GooDE and Bkan (13, p. 483) characterize the family Macrorham])hosidæ as: "Hemihran- 

 chiates with compressed body, armed with bony plates on belly and anterior parts of body". 



»0; p. 78 (40): 



L. Agassiz (la, Vol.1, p. 90) was the tirst to recognize the peculiarities of the scales. As 

 his description seems to have been totally overlooked by all later authors, 1 quote it here 

 in e.rlenso: "Les Bécasses de mer (Centriscus) ont un ty])e d'éeailles tout i)articulier. Ce sont 

 de petites esquilles lisses, cachées dans la peau, surmontées d'une tige cylindrique et courte 

 qui s'étale de nouveau à la surface de la i)eau en un écusson de forme trapézoidc. (^et 

 écusson montre plusieurs carènes qui rayonnent en arrière et qui se terminent par des pointes 

 assez eftilées". 



Later KNER(21b, p. 258(27)) pointed out the main feature of the scale, viz. that it is 

 composed of a basal part ("Wurzeln ") and a scale-plate ("P'lächenausbreitung der Schuppe")- 



O. HiiRTwiG has given a detailed description of their form and relation to the dermal 

 layers, including also some of the simpler forms, e. g. those of the eye, the base of the pec- 

 toral fin, and of the fin-rays. In addition, he has examined "eine zweite sehr kleine Art, den 

 Cenlriscus brevispiiiis", in which he finds a much simpler type of scale; hence he concludes: 

 "Alles macht den Eindruck, als ob das Hautskelet vom Centriscus brevispinis sich rückzu- 

 bilden im Begrilfe stände". According to Lütken, however, (24b, pp.586, 610), this "species" is 

 a young stage of Cgracilis Lowe*; Ilertwig's description therefore, and his figure (PI. 1, fig. 26) 

 do not concern stages of reduction but stages of development of the scales, such as will be 

 found also in the other species of the genus. 



L. Vaillant (33a, p. 126 and 33 b, p. 338) describes the scales anew and gives one figure 

 (PI. XXVII, fig. 3); apparently without knowing Hemtwios much more detailed description. The 

 plates as well as the difTerent forms of the scales arc not mentioned. 



11; )).82 (44): 



Information regarding the inner skeleton is given by Rosenthal (27), Agassi/ (1 b), Meckel 

 (15), Brühl (5a and b), Cünthkr (14a) and \V. Sorensen (32J; and in recent years by Stahks (30) 

 and SiEBENRocK (29) for the pectoral girdle and some ijoints in the skull. Apart from Rosenthal's 

 statement (quoted above under 9) regarding the elongation of the 4 anterior vertebræ, Agassiz 

 (I.e., p. 272) was the first, so far as I am aware, to give any information regarding the spinal 

 column and the interspinous bones. He states that there are 9 abdominal and 14 caudal 

 vertebræ. He further says, regarding the anterior vertebræ: 



"Les cinq premières vertèbres abdominales sont remarquables en ce que leur corps est 

 très-allongé, saillant en forme de double cône dans la cavité abdominale, et que les deuxième 

 troisième et quatrième ont de très-grosses et larges apophyses transverses qui s'étendent 

 horizontalement jusqu'à la plaque osseuse qui va de l'humérus au premier rayon de la dor- 



* Lütken's statement (24 b, p. 586(178)), that young individuals of C. gracilis of 17 mm. and 

 below are without ventral fins, is incorrect. I find the ventrals quite conspicuous, witli the fin-rays 

 discernible, in the smallest specimens of 7—8 mm. length. They might also perhaps be found (e. g. 

 through suitable staining) in stages of C. scolopax, corresponding to those of 10 mm. length, figured by 

 Emery (8, PI. I, fig. 12) and said by him to lack ventrals (8, p. 12.) 



