RECENT HYDROBIOLOGICAL INVESTIGATIONS 613 



the winter and early spring in order to reproduce in the shallow 

 relatively fresh, and cold water covering the coastal banks and 

 slopes. Possibly the shoaling movements of the herring are 

 reactions to such hydrographic changes. No doubt there are 

 herring in the North Sea off the east coasts of Great Britain 

 during the entire year, for it would appear to be the case that 

 the various herring caught there are local in their habitat. 

 But the fish only shoal, or become segregated, at a certain time 

 in the year, and it is only then that a fishery becomes possible. 

 (3) There are small local movements of fishes which lead to an 

 occasional abundance in certain restricted parts of the sea, and 

 in many cases these strictly local migrations are caused by a 

 local abundance of some animal on which the fishes feed. 

 Bottom-living lamellibranchs, for instance, such as the mussel, 

 are a favourite food of species of flat-fishes, and the distribu- 

 tion of these molluscs is variable from time to time in any 

 one place. 



It is always very difficult clearly to demonstrate the con- 

 nection between the abundance of fishes in a restricted sea-area, 

 and some set of physical conditions in the sea. This is partly 

 due to the difficulty of obtaining full and accurate information 

 with regard to the hydrographic changes, and also with regard 

 to the abundance of fishes. The periodic hydrographic cruises 

 of the last half-dozen years are now enabling us to supply the 

 first kind of information. These cruises are made every three 

 months, or, at the most often, once a month, but the develop- 

 ment of methods of interpolation, such as, for instance, the 

 use of the sine-function in obtaining values of the sea-tem- 

 perature for intervals intermediate between the dates of the 

 observations, enables us to make a continuous picture of the 

 state of the sea. Measures of the abundance of fishes in a 

 part of the sea are obtained from the commercial statistics of 

 fish landed from the fishing-grounds ; and during late years 

 the methods of collection of the fishery statistics in Great 

 Britain, and other North European countries, has been greatly 

 improved. As a rule the more perfect these methods of 

 acquiring information become, the closer is the connection 

 between biological and physical events in the sea. Neverthe- 

 less, there is great need of strictly scientific methods of 

 investigation with regard to the distribution and rate of 

 metabolism of the marine fishes. 



