SCALE OBSERVATIONS OF THE SQUETEAGUE AND PIGFISH. 299 



Kuntzmaiiu's (1824) classification of scales, while artificial and crude, was far in 

 advance of his time. It follows: 



1. Membranous scales — those with concentric lines. 



2. Semimembranous — membranous posterior field, but anterior field faintly marked, as in Clupea. 



3. Simple scales — no radii or circuli; simple center. 



4. Scales with a design. 



5. Scales divided into regions. 



6. Scales with prickles. 



7. Spinous scales. 



Agassiz (1834) gave great impetus to scale classification. He originated the four 

 groups — ganoid, placoid, cycloid, and ctenoid. His system was abandoned on account 

 of the great variability, but attempts are being made, it seems, to revive it. Cockerel! 

 and Miss Esdaile are working in this direction. 



Mandl (1839) claimed to have found definite characteristics for each family and 

 expressed his belief in their usefulness for distinguishing genera, and even species. 

 Peters (1841) repudiated this statement when he found both cycloid and ctenoid scales 

 in the same fish. Vogt (1842) was able to distinguish the different orders of ganoids 

 by their scales. 



Baudelot (18713) concludes that none of the characters can form a basis of classifi- 

 cation, since the presence or absence of spines — the most important scale character — 

 is too variable. Although the characters alone are of little value, yet taken together 

 they ought not to be neglected in forming natural groups. Tims (1906) distinguishes 

 the different groups of the Gadidse but goes no further into classification. 



The work of Cockerell (1910, 1911, 1913, 1915) and Cockerell and Callaway (1909) 

 on classification is more elaborate than that of any other recent investigators. Cockerell 

 says (191 1): 



It has been possible to test rather thoroughly tlie value of scale characters and the result has been 

 to show that while they are not rarely deceptive through convergence, they are, on the whole, of great 

 taxonomic importance. 



As indexes of classification he uses size, shape, spines, radii, and circuli. 



AGE DETERMINATION. 



The more important means of age determination is based on Steenstrup's (1861) 

 observation that all scales except placoid grow throughout life proportionately to the 

 size of the fish. Agassiz, however, believed that scales are laminate and that one 

 lamina was added each year. Baudelot also took this view, with slight modifications. 



By polarized light Carlet (1878) was able to distinguish old scales from young ones, 

 the former being birefringent, while the latter were monorefringent. Further, by 

 means of picrocarmine stain he was able to distinguish the newer lamina from the 

 older ones, the uncalcified parts staining red, the calcified parts staining yellow. 



Hoffbauer (1898, 1900) by observations on the hibernating habit of the carp, showed 



the supposition that the circuli were lamina edges was incorrect, since the number of 



laminae is not the same as that of the circuli and the number of laminae is greater than the 



number of years the fish has lived. He says that the number of circuli between annuli on 



1 



