584 MR. J. A. MILNE ON THE 



salmon will be examined upon the chance of finding one. Much 

 can be done with an ordinary high-powered pocket-lens, and for 

 close obsei-vation only a very low-powered microscope is either 

 necessary or desirable. I find a l|-inch objective powerful 

 enough for all purposes. One wants to be able to see as much as 

 possible of the scale at the same moment. 



Mr. Knut Dahl (' Age and Growth of Salmon and Trout in 

 Norway,' London, 1912) has shown that the size of a fish at 

 any period of its life may be deduced from its scales. 



A fish does not change its scales. From the time they form 

 they are retained through life unless removed accidentally. If a 

 fish does by chance lose one it is replaced by another of the same 

 size and shape but without the coi:icentric lines — see text-fig. 101. 

 It follows that as the fish grows, but remains covered by the same 

 number of scales occupying the same relative positions, the scales 

 must grow with the fish, and the growth of each scale be pro- 

 portionate to the growth of the fish. If, then, the protected 

 part of the scale be measured along its length from the centre to 

 the anterior edge and again from the centre (by which I mean 

 the nucleus) to, say, the first winter ba,nd, the lengths of the 

 scale at these two points will be proportionate to the lengths of 

 the fish at corresponding ages. An example will make this more 

 clear. Suppose we have a salmon 75 cm. long and that one of 

 its scales, when magnified, measures 55 mm. from the centre to 

 the anterior edge, and 1 1 mm. from the centre to the point at 

 which the rapid growth, consequent on migration to salt water, 

 is seen to begin. Then as 55 is to 11 so is 75 cm. to the length 

 of the fish when it entered the sea. The smolt was therefore 

 15 cm. (or just under 6 inches) long. Another scale from the 

 same fish magnified to the same degree, might be only 45 mm. 

 long, but then the other measure would be found to be 9, and the 

 sum would work out just the same. 



Dahl has proved his theory by measuring the scales of 

 hundreds of salmon and trout from different rivers. He has 

 always found that the actual average lengths of the fish in 

 difierent rivers at various ages, agree almost exactly with the 

 lengths calculated by him from the scales of the older fish. 



I have been trying to apply this method of measurement to 

 the scales of Pacific salmon, and I will give the results when I 

 come to deal with each species separately. First, however, 

 I must make some criticisms on what I have set out above as 

 my understanding of the claims made by Dahl. I tried to 

 check his theory on the scales of two salmon which had been 

 marked as kelts and subsequently captured as clean salmon. 

 Photographs of both of these scales form illustrations to 

 Mr. Johnston's second paper on salmon scales in the 25th 

 Annual Report of the Fishery Board for Scotland. The lengths, 

 on the occasion of each capture, are given elsewhere in the 

 24th and 25th Reports, so it is known that the salmon marked 

 No. 9194 was 27 inches long when taken as a kelt, and 31 inches 



