864 



SCIENCE. 



[N. S. Vol. XXII. No. 574. 



mechanism of color changes in fishes. We 

 are told (I., p. 8) that changes in the posi- 

 tion of the scales produce color changes ; 

 ' when they rise a little the color seems to 

 change.' ' Fine lines and concentric striae on 

 the scales ' are said to produce a bluish luster, 

 although there is 'no real blue pigment' 

 under them (I., p. 9). Again (I., p. 226) 

 ' certain metallic shades, silvery blue or iri- 

 descent, 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.' And yet there are several ref- 

 erences to blue pigment (I., pp. 8, 9, 129, 155) . 

 Pouchet^ showed more than thirty years ago 

 that the blue colors of fishes were due, not to 

 the structure of the scales, but to a layer of 

 cells containing guanin crystals, and to these 

 cells he gave the name of iridocytes. Cun- 

 ningham and McMunn,* though differing from 

 Pouchet in details, reached the same conclu- 

 sion. The pigments of fishes are lipochromes 

 (red, orange, yellow) and melanin (black), not 

 blue. All this is accurately and fully set 

 forth in the ' Cambridge Natural History.' 

 Moreover, Camichel and MandauF have re- 

 cently made the general statement that in 

 vertebrates all blue skins contain no pigment 

 but black, and all green skins no pigment but 

 black and yellow. 



These are, to be sure, in a sense errors of 

 omission, but there are other errors not of 

 this sort. Thus we are told (I., p. 91) that the 

 labyrinthine apparatus of Anabas scandens is 

 a ' contrivance for holding water above the 

 gills,' though Zograff° showed the apparatus to 

 be for the purpose of holding air for respira- 

 tion. Here again the ' Cambridge Natural 

 History ' would have furnished accurate in- 

 formation. On page 107 arterial bulb and 

 conus arteriosus are used as synonyms. One 

 of the strangest errors is in the statement (I., 



'Pouchet, G., Comptes Rendus, LXXIV., 1341- 

 1,343 (1872). 



* Cunningham and McMunn, Phil. Trans. Roy. 

 Soo. Lond., Vol. 184B, p. 765. 



^ Camichel and Mandaul, Comptes Rendus, 

 cm., 826 (1901). 



''Zograff, Quar. Jour. Micr. Sci., XXVIII., 501 

 (1888). 



p. 118) that in the deep sea form Ipnops ' the 

 eyes are spread out to cover the whole upper 

 surface of the head, being modified as luminous 

 organs.' This was Moseley's original opinion, 

 but, as he states in his final paper,'' it arose 

 from the fact that among the sections of 

 Ipnops given him for examination there was 

 an unlabeled slide ' containing a section of 

 the retina of some ordinary teleostean fish.' 

 His final examination showed him that the 

 ' peculiar organs have in reality no connection 

 with organs of vision. The eyes as well as 

 the optic nerves are entirely aborted in the 

 fish.' 



These instances from the general part of 

 the book show that it can not be uniformly 

 depended upon to present a full, accurate and 

 modern account. On the other hand, the 

 chapters on distribution, classification and the 

 history of ichthyology do not appear to err in 

 any of these directions. These subjects are 

 treated with so great fullness and knowledge 

 that nothing equal to them is to be found else- 

 where. The chapters on distribution have 

 the rare merit of relating the distribution to 

 the topography and recent geological history 

 of the regions discussed and are to be com- 

 mended for their wise conservatism. The 

 chapters dealing with the external structural 

 features and that on paleontological history, 

 though less technical, are usually clear and 

 well suited to various classes of readers. 



The descriptive portion of the book (some- 

 what more than one half) follows the method 

 long familiar to American zoologists in Pack- 

 ard's ' Guide to the Study of Insects.' The 

 salient characters of each family or larger 

 group are given and are followed by a running 

 and often gossipy comment on its commoner 

 or more interesting representatives. There 

 are no keys or specific descriptions, so that for 

 determining species President Jordan's other 

 books should be used. There is much differ- 

 ence of opinion among naturalists as to the 

 value of this mode of treatment. While it 

 affords much of interest to the technical stu- 

 dent, it commonly fails him when he is in 

 search of information about particular species. 



' Moseley, ' Challenger Reports,' XXII., 269. 



