Fakran — Seventh Reporl, on the Fishes of the Irish Atlantic Slope. 187 



A detailed examination of the scales of some species of Bathi/r/adus 

 (Gadovius) has been made by Gilbert and Hubbs (1920), and they reacli the 

 conclusion that, in the shallow-water foiins (ca. 200-o00 fathoms), rings which 

 probably represent annual periods of growth are clearly present, but that these 

 markings are very obscure or entirely absent in species inhabiting great depths. 

 These rings in the species examined were due to (a) a crowding together of 

 the striae (circiili), a form of ring not found in any of the Macrurids which I 

 have examined, or (5) a sudden bending of the stiiae along a line parallel to 

 the margin of the scale — a type of marking which corresponds very nearly to 

 that described above. 



The observations on the Macrurid scales show us : — 



I. That very few scales show rings clearly marked all round. 



?. On the same scale the same ring gives slightly different proportional 

 measurements along different radii of the scale, somelimes being nearer 

 proportionately to the lateral margin, but more often to the anterior. 



3. In the same fish different scales give different numbers of rings at 

 different proportional distances from the centre ; but usually a majority of the 

 scales will be found to agree in the relative positions of one or more rings. 



In the case of the herring, in which the theory of annual rings has been 

 most fully studied, it is found that : — 



1. Almost all the scales show rings clearly marked all round. 



2. On the same scale the rings are proportionately the same distance from 

 the centre on every radius. 



3. On the same fish almost all the scales have the same number of rings 

 at the same or slightly different distances from the centre, with the following 

 exceptions (Paget, 1920) : — 



{a) Individual fish may sometimes be found with a secondary or false 

 ring intercalated between two annual lings. 



(h) On the same fish individual scales may sometimes show a ring 

 which is not represented on the other scales of the fish. 



These two exceptions are not so frequent as to put any serious difficulty 

 in the way of determining a fish's age and rate of growth from its scales, at 

 any rate in its earlier years. 



In the case of the Macrurids here considered, the most reasonable 

 explanation of the condition observed seems to be that rings are formed at 

 periodic intervals on the scales, but that the two exceptions a and h have 



