29 



B. HELLAND-HANSEN 



according to the first method, and the II group according to the last method as 

 30-5 cm., or a difference of only o'g cm. as against the difference of 3'3 cm. recorded 

 in the table opposite group III. 



It may be of course that the values obtained by the scale-test are somewhat 

 high; and it must be owned that it was only found possible to subject 752 out 

 of the 1746 individuals in the two catches to the scale-test, and that it is chiefly 

 in the case of the smaller individuals that no examination could be made. 

 However not even this would be sufficient to account for the great discrepancy. 



In fig. 15 the results of the scale- tests for this section, G, are indicated by 

 open rectangles. If the high average values of length are somewhat reduced, 

 especially in the case of groups I and II, 

 they will correspond fairly well with the 

 growth-curve for section D (the unbroken 

 line). The broken curve in the figure 

 shows a considerably slower growth for 

 most of group II and for the whole of 

 group III than is recorded by the other 

 curve. It may be imagined that the 

 length-groups, shown in the figure under 

 age-group II, really belong to I ; and that 

 these shown under group III ought to 

 belong to group II; in which case the 

 curve for G would show a stronger 

 growth than what we have found in area 

 D. But if this is so, there must have 

 been in a large portion of the catches 

 and in practically the same locality two 

 separate lots of haddock both belonging 

 to the same year-group' though with 

 very dissimilar growth. This will become 

 clearer if we examine fig. 16, where we 

 see the size-distribution in a catch made 

 on 23rd October 1906. The groups are 

 not so absolutely differentiated as in some 

 other cases; the distance between them is less than usual. The first group in 

 the figure corresponds with the I group of the growth-curve; the second group 

 accords equally well both with the I group as per the scale-test and with the 

 II group of the growth-curve. 



We find something similar at the Dogger Bank (areaH); and Damas, after scale- 

 testing a catch made at the beginning of September 1906, has obtained the following: — 



LengtJi at the beginning of 



September 1906 Number of individuals 



Fig. 16. 



UELMÛTM IM CEJNTIMELTERS 



Length-groups in a catch of haddock^ 

 from area G. October 1906. 



» March has again been taken as the period when the y ear- groups commence. 



