AGE DETERMINATION FROM SCALES OF LAKE TROUT 



23 



Fkure 15. — (A) Scale of a lake trout 8.6 inches long, marked in September 1945 and recovered October 8, 1947. The 

 band of new growth is wider than the entire growth zone of the previous year. (B) Scale of a lake trout 10.8 inches long 

 with a right pectoral fin missing; no regeneration. Caught November 4, 1948. Only 3 checks were found on the scales. 

 As lake trout with the right pectoral fin removed were released in September 1945, the scales shovild have had 4 checks, 

 3 antnili, and 0-mark, if the fish were one of those marked. The missing fin was, therefore, abnormal. 



scales of some lake trout by the latter part of 

 Marcli, it could not be seen on otliers initil August. 

 The period for the start of new growth, therefore, 

 extends tlu'ough .5 months. Possibly, tiie period 

 wotdd he shorter for groups of fish, all caiigiit 

 from a small, localized area. The present collec- 

 tion of marked lake trout came from contiguous 

 but relatively extensive areas in the nortlieastern 

 part of Lake Michigan. A diversity of environ- 

 mental conditions in various localities, about 

 which there is at ijresent very little information, 

 may cause growth on the scales of local groups of 

 :<78:):if! o— ,^i(; —4 



lake trout to begin at different times so that when 

 the groups are combiiu-d, as in the present study, 

 tlie semblance of a long period for tiie begimiing 

 of growth would result. 



Assuming normal distribution, the combinetl 

 data fit, within the confidence limits at the 5- 

 percent level of probability, a normal cimaulative 

 curve witii tlie (7 = 20 days and the .'iO-percent 

 level on .Time 18. The test used for goodness of 

 fit was tiie Kolmogorov-Smirnov test described 

 by Massey (1951). Whereas tlie .")0-percent level 

 of the theoretical normal po|)uhuion falls on 



