170 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



those relating to the work of 1913, give the results of this 

 summary, and in the following sections whatever of 

 interest appears to emerge from a broad consideration of 

 the data available is discussed. 



(2) The Statistical Methods Employed. 



The actual measurements made at sea on board the 

 steamer and the police cutters are, so far as my own 

 observation goes, reliable and careful. The fish have 

 been measured on a board, into which a metre scale was 

 sunk flush with the surface. The head of the fish is 

 placed lightly against a bar at the zero of the scale, and 

 the centimetre division into which the extremity of the 

 tail falls is recorded. For some time a measuring board 

 was employed into which pins were stuck opposite the tail 

 of the fish. By counting the number of pins in each 

 centimetre division the length frequencies were read off. 

 It was, however, found to be quicker for one man to 

 measure the fishes, while another made ticks in a book 

 ruled oif in centimetre columns. In this way a catch of 

 even 2,000 plaice was very quickly dealt with. The fish 

 were always measured as soon as caught, so that no 

 correction is required to compensate for post-mortem 

 shrinkage. This correction was obtained by re-measuring 

 samples in the laboratory that had already been measured 

 on board ship 24 hours previously, but it was found to be 

 so small that it has been thought unnecessary to apply it. 



Grouping of the measurements has been avoided in the 

 working-up of the results. The arithmetical labour is 

 lengthened, because of the absence of grouping, but the 

 results are more accurate and have greater permanent 

 value on this account. In the tables given in each year's 

 report the results are collected for each month and fishing 

 ground; and in the general summary which follows, the 



