220 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



In this diagram the four vertical lines denote the 

 four dates of the investigation. The unbroken curve 

 denotes the average number of cod larvae per 5-min. 

 tow-nettings at 1 m. depth and 250 metres length, calcu- 

 lated by integration of the figures gained by the 20 tow- 

 nettings, which were made on each occasion at the four 

 stations in the fiord. The broken curve shows the 

 alterations in salinity at 20 m. depth in the middle station. 

 The dotted curve denotes the distance under the surface, 

 where, on the same station, the specific gravity of the 

 waier in situ* was 1,021, a value the importance of which 

 has previously been mentioned. 



If we study this diagram we shall observe that the 

 decrease in the number of larvae closely follows the 

 upheaval of the salt water and the rise of the isopykn of 

 1,021 towards the surface, expelling the " fry water " of 

 lower salinity. When the salinity again was lowered 

 and the isopykn of 1,021 again subsided the number of 

 larvae rose. 



Similar phenomena were observed also in the 

 Sandnesfiord, and the circulation in this fiord being 

 quicker, the alterations in the numbers of fry t were here 

 still greater. Also a study of the occurrence of the 

 Pleuronectid larvae shows exactly the same great 

 decrease in their number parallel to the expulsion of the 

 upper water-layer in which, they lived. 



These facts seem to me to indicate an intimate con- 

 nection between the movements of the water and the 

 movements of the eggs, larvae and fry. There is evidently 

 a more or less constant circulation of pelagic fry con- 

 nected with the circulation of the waters. The pelagic 

 fry evidently drift in and out according to the changing 



* The weight of the water in grams per c.cm. 

 f Also the number of eggs decreased more rapidly in this fiord. 



