METHODS OF INVESTIGATION. 5 



the second winter, and which show a repetition of the characteristics of the 

 first year's growth. I examined altogether ten of these parr, and the 

 scales of all showed growth-zones and winter-bands similar to those 

 illustrated. 



Further in the examples I have examined of parr and smolts which 

 have grown under natural conditions, and which were caught in spring, 

 summer, and autumn, I have found no difficulty in distinguishing the 

 appearance of the scales at the various seasons of the year. 



FlG. 4. Parr 12 cm. Softeland May 25, 1909 

 (magnified). 



FIG. 5. Parr 10 cm. Softeland June 11, 1908 



(magnified). 



The three next illustrations (Figs. 4, 5, and 6) depict the scales of parr 

 caught in the Softeland Eiver in the Os district, in May, June, and October. 

 They show clearly how the summer's growth is indicated on the scales. For 

 example, in Fig. 4 we see that the winter-band, formed during the second 

 year of the parr's existence, lies close to the edge of the scale and that only 

 two new broad rings, denoting summer growth, had been formed. The June 

 scale (Fig. 5) shows that these broad summer rings have increased in 

 number, whereas the October example (Fig. 6) clearly indicates the fact that 

 the formation of broad summer rings had ceased, and the narrow branching 



