- 150 — 



[Molva byrkelange) ; but we have as yet too little information to be able to say whether 

 this rule has absolute applicability. On the other hand it is certain, that the range of 

 temperature within which spawning takes place is everywhere, often much, 

 narrower than that within which the species may occur as a whole. A striking 

 example of this will be shown later, namely, the haddock (G. œglefinus), see p. 151— 52; 

 but it will be found that something similar, though perhaps not always to such a marked 

 extent, holds good for all our other species. 



If we examine into the conditions for the species, for which both the minimum and 

 maximum temperature for the spawning could be determined, we find that there is one 

 temperature at which spawning takes place to a larger extent than at the two extreme 

 points. As example m'ay be cited the case of the common ling {M. molva), which has a 

 minimum temperature of ca. 6° for the spawning and a maximum of ca. 10°, but which 

 spawns relatively little at these extreme points. On the other hand, much more spawning 

 occurs at a temperature of about 8° , which may therefore in this regard be called the 

 optimum temperature for the spawning of the ling. In the same way the other species 

 seem, so far as the observations go, to have an optimum temperature for their spawning, 

 which naturally lies as a rule in the neighbourhood of the average between the minimum 

 and maximum temperatures. 



The subdivision of the Atlantic species p. 148 into two groups according to the temp- 

 erature ("northern" and "southern" groups) naturally does not mean that all the separate 

 species within these are quite alike in their requirements. On the contrary, the data 

 given in the list p. 139 — 47 show that all gradations occur within the two groups, so that for 

 example in the "southern" group there are species which come near most distinctly north- 

 ern species of the other group (e. g. pollack [G. poUachius] and poor cod [G. minutus]) and 

 the reverse. 



Similar considerations to these regarding the temperature will undoubtedly prove ap- 

 plicable with regard to the other outer factors which exert their influence on the fishes. 

 It will thus certainly prove that each species has an optimum of depth and salinity for 

 the spawning, but the available observations are still too incomplete in these directions to 

 permit me to enter further into their discussion at present. 



G. Some biological conclusions 



From the records of observations given in previous pages it is possible to draw some 

 conclusions of a more general nature. 1 exclude here all mention of the arctic Gadus 

 saida and refer only to the remaining Atlantic species. 



The spawning region of a species is not identical with its distribution 

 as a whole. 



It was shown in the introduction that knowledge of the spawning region of a 

 species was of much greater importance to the understanding of its geography than 

 mere knowledge of where it occurs. The common freshwater eel [Angiiilla vulgaris) was 

 mentioned as a specially striking example of this, as it occurs almost everywhere in Eu- 

 rope in quite shallow fresh or brackish water, but in order to spawn must migrate out into 



