SEA-FISHERIES LABORATORY. 115 



It will be seen that the series of hauls is far from 

 being complete with regard to each area. This is partly 

 due to the fact that there are sometimes comparatively 

 few plaice on the grounds in question; but it is more 

 often due to difficulty in getting the Fishery Officers to 

 understand their instructions. This has really been the 

 most troublesome and discouraging part of the whole 

 investigation, and one feels that results of value may have 

 been lost by the failure to obtain an uniform series of 

 data for each of the fishing grounds. 



Only the actual data themselves are tabulated, and 

 the probable errors of the modes, etc., have not so far 

 been considered. It is quite clear that separate series of 

 figures representing the observations of several years, or 

 of several different areas, cannot properly be compared 

 with each other, unless due regard is paid to the statistical 

 errors of the values compared. But the necessity for the 

 calculation of these factors does not yet arise, since no 

 comparisons are made, except a few which are purely 

 provisional in their nature. It must, however, be con- 

 tinually borne in mind, whenever any two curves, or 

 series of figures, are compared with each other, that every 

 point on the curve, or figure in the series, is only an 

 approximation to the (unknown) real value, and is subject 

 to more or less error. It is always possible to find the 

 limits of this error, and when one makes a comparison, 

 the values compared must obviously be greater than the 

 limits of error. 



We may now consider the various fishing grounds in 

 detail. The figures relating to the Firth of Clyde, Luce 

 Bay, and the Bahama Bank area, are records of very few 

 hauls, and are only recorded for comparison with the 

 Lancashire and Welsh grounds. The Clyde off -shore 

 from Corsewall Point is a fishing ground characterised by 



