23g DISCOVERY REPORTS 



the coast and seem likely to affect the fishing over a much wider area. This raises the question whether 

 fish or sea-lions are the greater asset. While trawling on rough ground near the Falkland Islands would 

 be difficult, so that here the sea-lion would be the more remunerative, it is possible that if their numbers 

 were reduced fish farther afield might be turned to still greater account.' It is now known that sea- 

 lions feed largely upon cephalopods (Hamilton, 1934), but there is little doubt that they eat con- 

 siderable quantities of fish also. The cephalopods are themselves among the most important fish foods, 

 and the way in which sea-lions frequented areas in the open sea where fish were shoaling was remarked 

 by all who took part in the trawling surveys. The Falkland Islands are also the breeding grounds of 

 vast numbers of oceanic sea birds, many of which eat fish whenever they can get them. These, too, 

 probably play their part in keeping down the local fish population. 



Gunther's notes also include some pertinent statements about the invertebrate fauna of the region : 

 'Lithodes and squids are of direct economic importance to the South American market; macro- 

 planktonic forms like Parathemisto and Munida spp. are among the important constituents of fish- 

 food ; the rich benthos appears to reflect the character of the sea floor, and is thus bound up with the 

 distribution of fish. Very full notes upon the quantitative distribution of invertebrates were therefore 

 made, but it will be impossible to make full use of them until systematists have revised the taxonomy 

 of the several groups.' 'The plankton of the area appeared to be poorer than that of corresponding 

 latitudes off the west coast of South America, and on the sub-Antarctic whaling grounds.' 

 Unfortunately, one has to add that, owing to the war, most of the collections, both of plankton and of 

 benthos, still remain to be worked up. The magnitude of this task may best be judged from a quotation 

 from the summary of Gunther's unpublished report on the work of the third survey: 



'. . .the OTC was shot at 80 out of 131 Sts. and repeat hauls made at 24 Sts. BTS was shot at 

 II Sts., BNR at 46, oblique plankton nets at 50, and N50V at all Sts. Roughly 150,000 square miles 

 between the Patagonian Coast and the Falkland Islands, and from 44° to 54° S were examined. The 

 trawling stations were arranged in 11 lines and at a rough average were 50 mi. apart.' This report of 

 Gunther's includes memoranda upon visits to the fish markets of Montevideo and Buenos Aires, and 

 upon the history of the trawling industry off the River Plate, based upon the account of Devincenzi 

 (1926). Both have been of great help to me in writing this work. 



Throughout the period of these investigations and for some years afterwards, our collections were 

 supplemented by specimens of littoral fishes obtained at the Falkland Islands by the government 

 naturalists Mr A. G. Bennett and Dr J. E. Hamilton. Both these officers worked in close collaboration 

 with the Discovery Investigations; indeed, Dr Hamilton was seconded for service with us for many 

 years, his chief work being on the bionomics of the sea-lion, which happily gave opportunity for many 

 incidental observations on shore fishes. The value of their work, especially Mr Bennett's, can best be 

 appreciated from the frequent references to it in the main part of this paper and in Norman's (1937) 

 systematic report on the coast fishes. Norman's work gives us the taxonomic foundation without 

 which this could not have been written. In addition to the Discovery material he was aided by various 

 small collections made on the mainland coast by workers to whom he has already made acknowledge- 

 ment. Norman also had access to all the material from the region already preserved in the national 

 collection. 



TOPOGRAPHY OF THE SHELF 



The locality and extent of the region surveyed have already been described. The names of the features 

 most useful for general descriptive purposes are shown in the general chart of the area (Fig. 2). This 

 chart also shows the arbitrary division into northern, southern and intermediate regions best suited to 

 the handling of the data available. Most of the names are those of salient features of the coast and 



