"When members of any group of fishes became extinct, those appear to have been the first to perish which were the possessors of the greatest number of widely modified or _specialized_ structures. Those, for example, whose teeth were adapted for a particular kind of food, or whose motions were hampered by ponderous size or weighty armoring, were the first to perish in the struggle for existence; on the other hand, the forms that most nearly retained the ancestral or tribal characters--that is, those whose structures were in every way least extreme--were naturally the best fitted to survive. Thus _generalized_ fishes should be considered those of medium size, medium defenses, medium powers of progression, omnivorous feeding habits, and wide distribution, and these might be regarded as having provided the staples of survival in every branch of descent. "Aquatic living has not demanded wide divergence from the ancestral stem, and the divergent forms which may culminate in a profusion of families, genera, and species do not appear to be again productive of more generalized groups. In all lines of descent specialized forms do not appear to regain by regression or degeneration the potential characters of their ancestral condition. A generalized form is like potter's clay, plastic in the hands of nature, readily to be converted into a needed kind of cup or vase; but when thus specialized may never resume unaltered its ancestral condition: the clay survives; the cup perishes." (DEAN.) FOOTNOTES: [19] Günther, Introd. to the Study of Fishes, p. 192. [20] The cells which von Lendenfeld designates 'phosphorescent cells' have as their peculiar characteristic a large, oval, highly refracting body imbedded in the protoplasm of the larger end of the clavate cells. These cells have nothing in common with the structure of the cells of the firefly known to be phosphorescent in nature. In fact the true phosphorescent cells are more probably the 'gland-cells' found in ten of the twelve classes of organs which he describes. [21] See a more technical paper on this subject entitled "Relations of Temperature to Vertebræ among Fishes," published in the Proceedings of the United States National Museum for 1891, pp. 107-120. Still fuller details are given in a paper contained in the Wilder Quarter-Century Book, 1893. The substance is also included in Chapter VIII of foot-notes to Evolution: D. Appleton & Co. CHAPTER XIII THE COLORS OF FISHES =Pigmentation.=--The colors of fishes are in general produced by oil sacs or pigment cells beneath the epidermis or in some cases beneath the scales. Certain metallic shades, silvery blue or iridescent, are produced, not by actual pigment, but, as among insects, by the deflection of light from the polished skin or the striated surfaces of the scales. Certain fine striations give an iridescent appearance through the interference of light.