RECENT HYDROBIOLOGICAL INVESTIGATIONS 617 



good fishery year. The livers of the fishes were also small and 

 heavy ; and the condition of this organ is so good a test of the 

 general condition of the fish as regards nutrition that fishermen 

 always observe it. The year 1903 was, therefore, one in which 

 the late appearance of the shoals was associated with a state of 

 nutrition of the fish far below the normal ; and in that year the 

 temperature of the sea at Lofoten was higher than usual — that 

 is, it was a year of exceptionally strong Gulf Stream flow. 



The fishery statistics do, however, show that there is a 

 relationship between the physical conditions of the sea and the 

 biological conditions, though it is, of course, necessary to be 

 cautious and not to strain these data too much. The roes — that 

 is, the imperfectly matured ovaries of the cod caught — and the 

 livers are both important articles of commerce, and the quantities 

 of both products exported from Lofoten have been recorded for 

 many years by the Norwegian fishery authorities. The roes 

 are used as food, and the livers are made into oil. The figures 

 for the take of roe are rather imperfect, and, further, the total 

 quantity of roe taken depends on the proportion of fish which 

 have not yet spawned — a proportion which falls off as the season 

 progresses. The take of liver will be in direct proportion to 

 the take of fish. If, however, the quantities of roe and liver be 

 expressed as numbers of hectolitres per thousand fish caught, 

 we obtain a measure of the degree of nutrition of the fish, as 

 indicated by the grade of development of these organs. When 

 these ratios are calculated for a number of years, and compared 

 with the hydrographical conditions of the sea, it is seen that 

 minimal takes of roe and liver are to be associated with sea- 

 temperatures at Lofoten which are above the normal ; con- 

 versely, maximal takes of these commodities are associated with 

 sea-temperatures which are below the normal, It is not possible 

 to compare the sea-temperature, and the quantity of roe and 

 liver taken, for a long period ; but the air-temperature can be 

 used for this purpose, and the latter rises and falls with the 

 temperature of the sea at the surface in the neighbourhood of 

 the Lofoten Islands. 



Fig. 8 has been constructed by Helland-Hansen and Nansen 

 to show this parallelism. The air-temperature values are 

 expressed as " anomalies," that is deviations above and below 

 the mean temperature of a great number of years. These 

 values are represented by the thick line I. in the figure. The 



