L>-40 THK VARIATION, RACES AND MIGRATIONS OF THE MACKEREL. 



that of values below 27 ; whereas for each sample of about* 100 fish 

 from Ireland, the values above 27 are, with only one exception, in 

 excess of the values below that number. The constancy of this 

 relation for each of the geographical areas mentioned appears to me to 

 show that 100 is very nearly a sufficient sample, in the sense already 

 defined, so far as this character is concerned. 



As a matter of fact, the four important regions for mackerel fisheries 

 round our coasts, viz., the North Sea, the Channel, the South and West 

 coasts of Ireland, are each represented by a totality of fish of not less 

 than 300 ; but I have entered at some length into this question, because 

 I think that the division of the local groups into unit-samples provides 

 a valuable means of checking the significance which may be attributed 

 to the differences as determined for the local groups en masse. 



II. The Characters Investigated. 



After a preliminary study of the question, it seemed desirable to 

 select numerical rather than dimensional characters for investigation, 

 their variations being less dependent than those of the latter upon 

 variable factors, such as disproportionate growth, food -supply, and 

 enlargement of the reproductive organs. The characters eventually 

 selected were the following : — 



1. The Number of black Transverse Bars or Stripes across the sides 

 of the fish, beginning at the point where the lateral line meets the 

 posterior border of the scapular arch, and ending just behind the last 

 dorsal finlet. 



2. The Number of the same Transverse Bars which meet or cross the 

 lateral line. 



3. The Number of round black Dorso -lateral Intermediate Spots 

 situated between the Transverse Bars. 



4 The Number of Finrays in the First Dorsal Fin. 



5. The Number of Finrays in the Second Dorsal Fin. 



6. The Number of Dorsal Finlets. 



The condition of each fish in regard to these six characters was 

 recorded from the commencement of the investigation, in addition to 

 the length of the fish, and, in the majority of cases, its sex. But 

 I soon began to take note of other characters, in the hope of finding 

 them useful in the end, and among these were the following : — 



7. The general Shape and Regularity of Arrangement of the Trans- 



* The irregularity in the numl>er of fisli in the samples from Ireland was due, I under- 

 stand, to a curious system of depredation during transit which prevails in those parts, and 

 is sanctioned by custom or, at least, endurance. 



