148 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



It will be seen that the total numbers of hauls made in 

 each year vary greatly. For instance, 61 trials with a fish 

 trawl were made in 1909, but only 16 in 1905. Further, the 

 distribution of the trials throughout the various months of 

 the year is not the same in evjery year. These irregularities 

 lead to some degree of accidental variation, though they do 

 not matter so much as might be supposed. The object of 

 the " trials " was to " get fish," and so when the number of 

 hauls is small, the plaice are not very abundant. Also the 

 trials were generally made during the months when it was 

 known that plaice were present, and these months are not 

 always the same in different years. Generally then, the 

 results represent the abundance of plaice in the locality, 

 sampled better perhaps, than if they had been made rigidly 

 according to a pre- determined routine. 



Another irregularity is the variable duration of the hauls, 

 but the time occupied in each does not differ very greatly 

 from one hour. So although the " average catch per haul is 

 not quite the same thing as the " average catch per hour," the 

 difference is not great enough to confuse any conclusions 

 based on the assumption that these averages are the same. 

 The natural irregularities inseparable from the method of 

 sampling are also a source of trouble, so that a graph of the 

 average number of plaice caught per haul or hour's fishing, 

 shows fluctuations which greatly exaggerate the variability 

 from year to year. 



Thus two or three exceptionally large hauls in one year 

 raise the average, though inspection of the general run of the 

 figures may show that quite mediocre results were the rule. 



Therefore some method of expressing the general results, 

 and at the same time minimising the importance of such 

 exceptional hauls, must be adopted. This has been done, to 

 some extent at least, by combining the results for each year 

 with those of the one before and the one after ; that is to say, 



