-38- 



little owing to our not being able to base our calculations upon direct weigh- 

 ings. Still it would be vain to strive after absolute accuracy here: for we must 

 reflect that our material — some 70,000 haddocks distributed over a period of 

 four years and over vast areas — is far from being wide-reaching enough to 

 yield more than imperfect information as to the conditions really existing. 



In what follows then we will merely endeavour to ascertain if there are any 

 essential differences between the total weights at different times and places, and 



J3 •4-0 'f5 



LELrfSTH IM CCNTIMtTE.R5 



60 cm. 



Fig. 28. Average total catch of haddock by the research steamers "Huxley" and "Poseidon" during 



the years 1902 — 1906. Dotted curve represents the relative average number per trawl-hour for each 



centimeter-size, without regard to either time or place. The other curve shows similarly the relative 



average weight per trawl-hour for each centimeter-size 



we will try to find approximately what is the relation between the weights of 

 the different size-groups. To many of the questions D'Arcy Thompson's paper 

 supplies a far better answer than what we have been able to give from the 

 comparatively meagre material in our possession. 



In Fig. 28 we find once more a curve giving the same average number per 

 hour that was shown in Fig. 4; in this case it is denoted by a broken line. It 

 has been adopted to express weight by means of multiplying the number of the 

 different centimeter-sizes by their average weights, according to Fulton's table 

 in the XXIst Report of the Scottish Fishery Board (1902) Part III. The result 

 thus arrived at is reproduced in the completed line, which therefore shows the 



