— 24 — 



only) the difference is surprisingly great, averaging no less than 5 vertebrae. It is, 

 moreover, already noticeable in the outward appearance of the fish, the specimens 

 from the mouth of the fjord having a more elongated form than those from the 

 upper reaches of the same. 



This decrease in the number of vertebrae from the mouth of a fjord upwards 

 along its course has been observed at so many different places in our waters as to 

 leave no doubt that the phenomenon in question is one of regular and general 

 occurrence. We may therefore conclude, that the inner waters of a fjord 

 present certain »external conditions« having the effect of reducing 

 the number of vertebrae in Zoarces^; in other words, that the low values 

 noted for this character in the case of fjord samples are, in part at least, to be 

 regarded as a merely phsenotypical phenomenon, and in no wise indicative of 

 genotypic differences. Naturally, however, this does not exclude the possibility that 

 genotypic differences may exist between the separate stocks of Zoarces in our 

 waters. 



As to what external factor or factors in the inner waters of a fjord may be 

 responsible for this decrease in the number of vertebrae, nothing can at present be 

 said. The lower degree of salinity in the water is a feature generally considered 

 as significant in such respects; as a matter of fact however, an examination of the 

 average number of vertebrae in the numerous samples investigated, in comparison 

 with the figures for salinity at the localities concerned, furnishes no support for 

 this theory, samples with exactly the same average number of vertebrae being 

 found at places differing widely in regard to salinity. 



Moreover, the values for the other characters appear to be dependent upon 

 other laws than those which govern the variations in the number of vertebrae. 

 Thus the number of pectoral rays, for instance, is as a rule found to 

 increase from the mouth to the inner waters of a fjord; i. e. the reverse 

 of what was noted in the case of the vertebrae; this does not, however, apply to 

 all the samples investigated. 



Altogether, from a study of the mass of material already examined, it would 

 seem that we have here to deal with a problem of highly complicated character, 

 involving considerable difficulty, and hardly to be satisfactorily solved without 

 having recourse to experimental methods. It is therefore the more fortunate that 

 the species chosen should be one which, in contrast to the eel, is extraordinarily 

 localised and highlj' variable, with the further peculiarity of being viviparous, all 

 of which features would seem to render it especially suited to the purpose of 

 experimental investigation. 



' Strictly speaking, there is still a further possibility, that the lower figures for no. of vertebrae 

 noted in upper fjord waters may be due not to the direct effect of certain external conditions, but 

 to the indirect effects of same operating through selection. 



