THE VARIATION, RACES AND MIGRATIONS OF THE MACKEREL. 245 



will much improve about September." Placed in formaline on arrival 

 at the Laboratory, and examined Sept. 19th, 24tb, and 25th. 



(2.) Sept. 3rd, 1897. 99 fish. Caught with nets off the Old Head of 

 Kinsale. "A fair average specimen of the fish now being taken." 

 Forwarded by Mr. Carroll. Placed in formaline on arrival at the 

 Laboratory, and examined Sept. 16th to 19th. 



(3.) Sept. 17th, 1897. 92 fish. Caught with nets off the Old Head of 

 Kinsale. "A fair average specimen of the fish now being taken." 

 Mr. Carroll. 



(4.) July 1st, 1898. 100 fish. Mr. Carroll. 



Smerwick, March 12th, 1898. 99 fish from Smerwick Harbour, 

 County Kerry; forwarded by Mr. John McKenna, Butter Merchant, 

 Dingle. 



Brandon (1), April 16th, 1898. 45 fish. Caught at Brandon Creek, 

 County Kerry ; forwarded by Mr. McKenna. 46 fish actually arrived, 

 but the smallest (length, 12| inches) has been eliminated, owing to its 

 marked divergence in size from the rest of the sample. 



(2.) April 23rd, 1898. 101 fish. Caught at Brandon Creek; for- 

 warded by Mr. McKenna. 



IV. Size of the Mackerel Investigated. 



The length of the body has been taken as the basis for comparing 

 the relative sizes of the fish investigated, and this has been measured 

 in all cases from the tip of the closed jaws to the median extremity of 

 the fork of the tail. The longer rays of the tail-fin are subject to 

 accidents in capture and transit, and cannot therefore be included in an 

 accurate comparative table of measurements ; but it should be remem- 

 bered that these rays are generally included in ordinary measurements 

 of body-length, which would consequently be somewhat in excess of 

 my determinations. The fish have always been measured in the same 

 way, by placing them with their left sides uppermost upon a measuring- 

 board, with the tip of the jaws touching a vertical plate immoveably 

 fixed at the zero of the scale, which is accurately ruled to quarters 

 of an inch. The length of the fish has been read off to the nearest 

 quarter. 



The observed number of specimens of the different sizes represented 

 in each consignment of fish is recorded in table A. The general size of 

 the fish representing each locality can, however, be more readily 

 gathered from the more condensed statements given below, which are 

 in percentages of the total number from each locality. In Table I. the 

 size is indicated in inches, the fractional differences recorded in table A 



