SEA-FISHERIES LABORATORY. 169 



II. On the Methods of Trawling 

 Observations . 



Any attempt to study closely the changes taking place 

 in the fish population of an area by means of such a 

 series of observations as we refer to above, discovers so 

 many apparent irregularities that we are forced to con- 

 clude that only a very general conclusion can be drawn 

 from such data, and then only when averages founded on 

 considerable series are the values compared. It is very 

 rjrobable, for instance, that the five-yearly averages we 

 have given above do indicate with some accuracy the 

 condition of the Blackpool and Mersey grounds, but it is 

 just as probable that any attempt to extract more than 

 this from the observations, say the variation from year to 

 year, is open to objection. If the averages of all the 

 catches made in each year were compared with each other, 

 errors would probably be made, for the abundance of fish 

 in any place varies enormously from month to month 

 during the year, and the hauls have not always been 

 spread regularly over the year but are more numerous in 

 the summer and spring of some years, and in the autumn 

 of others. In comparing years with each other we are, 

 therefore, compelled to compare only the catches made 

 during the corresponding seasons, and in doing so we 

 reduce considerably the number of hauls on which to cal- 

 culate averages. Results based on only one haul per 

 month or quarter are open to still greater objection than 

 those deduced from small averages. 



The exact conditions affecting the number of fish caught 



in a trawl net, employed for experimental purposes, are 



therefore, worth considerable study and we do not know 



of any observations published* with this aim. The In- 



* Except perhaps some observations made by Mcintosh. See Report 

 of the Royal Commission on Trawling ; "On the effect of successive 

 hauls on the same area," p. 374. 



