LIFE HISTORY OF LAKE HERRING OF LAKE HURON 279 



5. Whether the growth of the fisli ui lengtli is retarded or arrested at the time of formation 

 of the annuli? 



6. What factors are responsible for tlie arrest of or retardation of growth in fish and scales? 

 (Van Oosten, 1923.) 



The last three questions I have attempted to answer in another place (Van 

 Oosten, 1923). It remains to discuss the first three questions. 



IDENTITY OF SCALES THROUGHOUT LIFE 



Were scales commonly or regularly shed and replaced by others they could not 

 be made use of in life-history studies. It is a tenet of the method that they retain 

 their identity throughout life, that only a few are lost accidentally and replaced. The 

 many life histories unraveled by the scale method are, in themselves perhaps, proof 

 enough that identity persists; but some of the well-established facts in proof of 

 identity are these: (1) That the nuclear area or central part of the scales of old fish 

 of a species is structurally identical with the scales of young fish (Snyder; see p. 300 

 following). (2) That regenerated scales, which replace those accidentally lost, have 

 a central portion of quite a different type from that of normal scales. (The characters 

 of regenerated scales are distinctive and generally easily recognized. Scott (1912) 

 and Creaser (1926) established this experimentally, and I have extracted normal 

 scales from carp and found them replaced by typical regenerated scales. I have also 

 found regenerated scales covering the repaired injuries of several herring.) (3) That 

 scales increase in size as long as the fish grows. 



CONSTANCY IN NUMBER OF SCALES THROUGHOUT LIFE 



If scales of typical teleosts, their number remaining constant, were in contact by 

 their edges they must grow in proportion to the body, else the surface would not be 

 always covered; the relation between scale length and body length would be mechani- 

 cal; but as these scales overlap, they may or may not grow in length in proportion 

 to the body's length. The question of whether they do or do not is one of physiologi- 

 cal growth correlation and may be answered only by observation. 



The number of scales in the lateral line is made use of in discriminating species, 

 and it is well known that it varies within the species. The question is whether the 

 individual differences in scale number arose when the scales were first laid down and 

 have continued since then, or whether the scales increase in number during the life 

 of the fish. Do the fish that were large at the time of scale formation develop more 

 scales at that time than smaller fish, and does their number remain constant? If 

 the number of scales (for example, in the lateral line), increases with age, they can not 

 grow in proportion to the body's growth, for it has been shown by Lea (1919) that 

 when a new scale is formed in place of one accidentally lost the growth rate of sur- 

 rounding scales is retarded. 



That the size of the scales varies with their number is shown by the fact that of 

 45 lake herring 240 millimeters in length, 25 had 80 or less (average = 77.9) scales 

 in their lateral line, while 20 had 81 or more (average = 82.8) scales in the lateral line. 

 The average length of the scales of the former group was found to be 5.16 milhmeters, 

 while that of the scales of the latter was 5.05 millimeters, or 0.11 millimeter less. 

 99760—29 2 



