HISTORY OF THE PACIFIC SALMON. 



581 



the edge of the scales at the time the salmon left the sea, and 

 this chickening is apparent both on the anterior and posterior 

 parts of the scales if the edges have not become too much 

 worn during the sta-y in fresh water, I account for it by 

 supposing that the materials which go to foi-m the scales 

 are still being secreted in fresh water, and that as the skin 



Text-fio-. 100. 



^ 



l^-ffr-fi- 





Scale of 33 lb. Saluiou {Salmo salar) caught in the Taj' in August 1903. 

 This fish had spawned twice, and the marks are shown at C and D. 



pockets have ceased to grow they can be deposited only on the 

 scale edges already formed. The cells from which the concentric 

 lines on the upper surface are evolved are already dead, except 

 close to the periphery of the scale (Klaatsch, ' Zur Morphologie 

 der Fischschuppen,' 1890, and Stuart Thompson, ' Journal of the 

 Marine Biological Assoc' vol. vii. no. 1, 1904). This in itself 



39* 



