116 DEPARTMENT OF THE NATAL SERTICE 



1. Samples from the waters about Prince Edward Island, i. e., west and 

 south of a line drawn from cape North, Cape Breton, to the eastern point of 

 Prince Edward island, and thence on to cape Gaspe in the province of Quebec. 



2. Samples from the waters about Magdalen islands. 



3. Samples from the coast of Newfoundland (all from the west coast, ex- 

 cepting one from White bay). 



4. Samples from the Atlantic side of Cape Breton island. Grand Narrows, 

 Ardoise, coast of Nova Scotia, Bay of Fundy, and from Gloucester, Mass. 



The following pages will be devoted to a description of the results obtained from 

 i.ge-determinations and growth measurements, arranged on the basis of the above 

 group-division. The first question to be dealt with will be, how far fluctuations occur 

 in the relative numerical value of the year-classes, and if such occur, whether they 

 become apparent in the same manner throughout the entire area embraced by the 

 samples, or whether they vary in the different waters. Thereafter, the growth in the 

 various localities will be discussed. And finally, by summing up the information 

 obtained from the study of the age and growth, and discussion of the material avail- 

 able as to number of vertebrae, etc., an attempt will be made to deal with the question 

 as to the possible existence of different races or tribes of herring in the Canadian 

 waters. 



XL— AGE. 



a. The tvaters about Prince Edward Island. 



From these waters we have fiv^e samples of grown fish, taken by gill-net at different 

 places in Northumberland strait; two of these are from May, 1914; the remaining 

 three from May, 1915. In addition to these, there are also two samples of herring 

 caught in traps in the neighbourhood of Souris during the summer of 1915, and eight 

 samples from drift-net catches made during the cruises of the drifter No. 33 at various 

 places in the gulf, in June and July, 1915. The chart, fig. 28, shows the localities. 



The trap-and drift-net catches contain a quantity of young specimens, and on 

 examining scales from these we find as will subsequently be shown, that the summer 

 growth for 1915 commenced some time in June in the case of the young fish, probably 

 during the second half of the month. It will therefore be reasonable to suppose that 

 the older and grown specimens in the samples from May had then not yet commenced 

 their new summer growth, and that, consequently, the last summer zone in the scales 

 would represent the growth of the fish during the summer of the year previous t^ 

 capture. 



The other samples, containing older fish, date from the time between the end 

 of June and the end of July. With regard to these therefore, we cannot be so sure as 

 to whether summer growth has commenced, or if so, whether it has commenced in all 

 cases. Some doubt arises in assigning the fish to definite year-classes (vide p. 119), 

 in addition to which the length of drift-net used by No. 33 was composed of greatly 

 varying widths of mesh, and thus rather calculated to take herring of all sizes than to 

 take them in the natural proportion between the different size-groups. The value, 

 therefore, of the samples in studying the proportions of the different year-classes is 

 somewhat difficult to determine, and the samples of grown fish from Northumberland 

 strait have consequently been kept apart from the rest. Fig. 29 shows the percentual 

 age-distribution in these five samples. 



It will immediately be noticed that the implements have here succeeded in 

 capturing fish of most widely differing age, from specimens 4 years old to individuals 

 of (apparently) some 20 years, a circumstance resembling that noted in the case of the 

 drift-net hauls of Norwegian grown herring, but in the present instance perhaps even 

 more marked. And although we here lack such means of gauging the value of the 



