18 



DEPARTMENT OF TEE NATAL SERVICE 



year class of the fii'st tj-pe and the five and three year classes of the last type were not 

 represented in the 1917 collection. The one year stream type predominated even more 

 largely than in 1916 on account of the lack of fish in the two year stream type, as the 

 percentage of sea type fish was much the same. In the one year stream type, the four 

 year fish made up even a larger percentage (95.4) of the whole number than in 1916 

 (82.9). There was little diflference in the percentage of males and females in the two 

 years, the females being slightly in the majority in each case. 



With the exception of the two year stream fish, the average length was slightly less 

 in all classes and types in 1917 than in 1916, the average difference being 0'-5 inches. 

 The growth in the first year was so consistently higher in the 1917 collection that it 

 would seem that the fish must have come from different parts of the Fraser river water- 

 shed in the two years. This difference is more than made up in the 1916 lot by the 

 more rapid growth in the second and third years. Which collection is more typical it 

 is hard to say but jwssibly since those caught in the Fraser river in 1917 agree with 

 those caught off Victoria and in Discovery passage, they are more likely to be typical. 

 Judging from the size of the yearlings this is likely to be the casa 



Looking at these figures there may be something in the contention that the fish of 

 what has been the largest year of the four year cycle, are smaller than in the other 

 years. 



Coho. 



The coho of the 1917 run, 1,1-17 in number, were obtained from localities within 

 the strait of Georgia and ad.iacent waters, and although they are considered in four 

 disti-act lots, most of them were caught before they were mature enough to indicate 

 to what river or creek they would have returned if they had been allowed to proceed 

 to the spawning grounds. Most of those obtained at the Quathiaski cannery (407) 

 were caught around Cape Mudge and from this point to Heriot bay. The Lasqueti 

 cannery specimens (417) were caught in the vicinity of Lasqueti and Tecxada islands. 

 The Fraser river specimens (89) obtained at the Na-aaimo cannery had become defi- 

 nitely localized on the way to the spawning grounds, but the remaining 504 obtained 

 at this cannery were widely distributed from Lasqueti island, Qualicum, Northwest 

 bay, Winchelsea islands, all the way to Gabriola pass and Oowichan gap (Porlier pass"). 

 <They were therefore a cosmopolitan lot but they could not well be kept distinct, 

 although differences were plainly to be seen at times, since the one day's catch was all 

 put together on the cannery floor. Even although there is not the same definiteness ' 

 in delimitation as in the case of the sockeye, there is still enough to make it wj^'th 

 while to draw comparison. 



All of the 1917 coho examined had gone down to the sea some time during the 

 second year and were in their third year when they were caught. Iti some cases where 

 the migration had been delayed until late in the second summer, the scale had much 

 the appearance of one from a fish that had spent over two years in fresh water, but as 

 the central portion of these scales corresponded with the complete scales of cohos in 

 their second year, caught in fresh water as late as the end of Ju-ne, the conclusion 

 that they could be fish in the fourth year, haKnng spent over two years in fresh water 

 can scarcely be justified. 



In all cases the females were more numerous tlian the males, the Quathiaski fish 

 showing the greatest difference, as the following table indicates: — 



The length shows much variation. 



