SEA-FISHERIES LABORATORY. 181 



artificial and misleading measures of dispersion. The 

 group of fish cut out from the whole distribution lies 

 almost entirely on the positive side of the mode (which is 

 evidently the natural origin). They are not the plaice 

 most characteristic of the population inhabiting this 

 ground that we thus segregate, but plaice which are on 

 the whole larger than the truly characteristic ones. If, 

 on the other hand, we take 50 % of all the fish which are 

 larger and smaller than the modal length, but as little 

 larger or as little smaller as possible we get a natural 

 measure of the dispersion, and a true notion of the kind of 

 fish that are characteristic of this fishing ground. Taking 

 either quartiles from the median, or the probable error 

 (as deduced from the standard deviation), we find that 

 the characteristic plaice population is one the length of 

 which is greater than about 19 cms., and less than 24 cms. 

 — obviously an inaccurate picture of the fish population. 

 But taking the grouping about the mode we find that the 

 really characteristic plaice which inhabit this ground are 

 those which are greater in length than about 17^ cms. 

 and less than 22 cms. Obviously the effect of regulations 

 (size-limits and the like) would be different as we 

 consider one or other of these measures of dispersion. 



This discussion seems rather an academic one, but it 

 is desirable that we should express the results of the 

 measurements made in the most natural manner. These 

 results are, it is claimed, of considerable " practical " 

 importance, for they are based on a large number of 

 measurements, and they represent fairly well the condition 

 of the Lancashire plaice grounds during the quinquennial 

 period 1909-1913. How valuable they may be for future 

 comparison may be appreciated when we consider how 

 valuable a similar series of measurements made 25 veins 

 ago, when the Committee first began its work of regulation, 



