CA\AIJ1A\ Fh-^HERIhJs EXI'EUITIOX, 191',-lo 163 



absolutely independent seleetiou, the task of choosing out specimens was delegated to 

 another party, who simply noted down twenty of the numbers which had been assigned 

 in the tables to the fish of the mentioned year-classes in the two samples. The sample 

 from Newfoundland was, moreover, the one with averages differing most from the 

 total average for all Newfoundland samples, the deviation tending in the direction 

 of the averages for the Magdalen islands. It may therefore be admitted that the 

 results obtained will not present any unjustifiably favourable view of the possibilities 

 inherent in such comparison. It was found that in seventeen out of twenty instances 

 the sum of the squares was the less in comparison with the averages for those fish to 

 which the specimen in question belong. The reverse was found in the case of one 

 specimen from the Magdalen islands, and two from Newfoundland, these differing 

 more from the averages of the fish whence they were taken. This result seems to 

 suggest that it should be possible, by means of the figures obtained from growth 

 measurements, to determine, in many cases at least, whether an old fish taken in the 

 gulf of St. Lawrence belongs in point of growth to the type found off the ;^[agdalen 

 islands, or to that found in the Newfoundland waters. In order to obtain the best 

 results, good averages would be required, i.e.. it would be necessary to examine large 

 samples, taken during the spawning season, and on the actual spawning grounds, such 

 choice of time and place being the best means of procuring " pure " samples, without 

 admixture of foreign elements. The comparison should be made not as has here been 

 done with averages for different classes (Magdalen islands 1903 and Newfoundland 

 190-i) but with averages for the same year-classes in samples from the different 

 spawning grounds. 



As to how far any similar individual distinction may properly be made with regard 

 to the herring in other Canadian waters, we cannot at present decide anything with 

 certainty, the material from the Atlantic areas being insufficient. iNfost probably, 

 however, we should in the majority of cases be able to determine whether a herring 

 tiken between Cape Breton and Newfoundland should be regarded as a Newfoundland 

 herring in the strict sense, or not. always pro\'ided that the investigations are so con- 

 tinued as to furnish a sufiScient fiuantity of material for computing the averages (with 

 standard deviations). 



LITERATURE REFERRED TO. 



I Dahl, Knut. The scales of the herring as a means of determining 

 age. growth and migration. Rep. on Norweg. Fishery and 



Marine Investig.. . . .^ Vol. II. No. 6. 1907 



II Dahl, Knut. The assessment of age and growth in Fish. Internat. 



Rev. d. gesamt. Hydrobiol. u Hydrograph Bd. II. H. 4-5. 



III Heincke, Fr. Naturgesehiehte des Ilerings. Abhandl. des Deutsch. 



Seefischerei-Vereins Bd. II, 1898 



IV Hellevaara. E. Undersiikningar rorande stromingen i sydviistra 



Finland. Finlands Fiskeriar Bd. I, 1912. 



V Hjort. Johan. Fluctuations in the Great Fisheries of Northern 

 Europe. Rapp. et Proc.-verb. du Cons. Internat. iwur I'Explor. 



de la Mer Vol. XX, 1914. 



VT Hjort, Johan. Investigations into the Natural History of the 

 Herring in the Atlantic Waters of Canada. 

 Suppt. 5th Ann. Rep. Dep. of Naval Service, Ottawa, 191.o. 

 VII Hjort. Johan and Lea, Einar. Some results of the Internat. 



Herring Investigations, 1907-1911. Pub. de Circonst.. No. 61, 1911. 

 VIII Hoffliauer, C. Die Altersbestimmung des Karpfens an seiner 



Schuppe. Allg. Fischereiztg .No. 19, 1899. 



IX Lea. Einar. On the methods used in herrinii- investigations. Publ. 



de circonst No. 53, 1910. 



