108 



DEPARTMENT OF THE NAVAL SERTICE 



6 GEORGE V, A. 1916 



Fish McurJcei, St. Andrews, August 8. 



Brandy Cove, August 12. 



Number 



373 



374 

 375 

 376 



Length. 



Cms. 

 8 

 7-8 

 6 2 



8 



To bring out the significance of the results contained in these tables, length- 

 frequency curves were plotted. At first all the fish were plotted on one large graph 

 as they were measured, but as^the season advanced and new fish were added to the 

 graph, it was found that the curve lost what little form it originally had. This was 

 doubtless due to the increase in length of the fish as the season advanced, throwing 

 them out of their classes. Hence this method was abandoned and only measurements 

 from catches, comparatively close together in time, were plotted on a single graph. 

 This curve was interesting in that the two catches were taken at practically the 

 same time, and the bulk of the fish fall within comparatively narrow limits in 

 regard to length; two shorter and several much longer fish were omitted from the- 

 graph. This curve seemed to show that the details in the contour of the curve are 

 meaningless, for the high places in one fill up the low places in the other, so that the 

 sum does not resemble the curve for either of the catches in detail. It appeared 

 also that most of the fish fall within the three-year class (the average length of fish 



