328 Fart III. — Twentieth Annual Report 



third net young long rough dabs, Norway pouts, and sprats were found 

 which had passed through the meshes of the inner net. The inside of 

 the end of the fine-meshed net was also lined with mosquito netting, 

 somewhat larger in the mesh than that represented in the figure (r), but 

 when used on the bottom it was ruptured. 



For enveloping the cod-end of the trawl, I have lately tried a cotton 

 net with meshes as represented in B, viz., about 6 mm. {\ inch) square. 

 It has been used on three occasions on the bottom, without mishap, and 

 by means of it smaller fishes have been taken. I am of opinion that a 

 net with meshes of even 4 mm. (y^^ inch) could be used in this way on 

 the otter-trawl with safety, provided the arrangements above described 

 are attended to, viz., the employment of a special lightly-made hempen 

 cod-end, with one-inch meshes, the buoying of the cod-end, and par- 

 ticularly when working on or near hard or rough ground, the use of 

 a canvas chafer under the whole of the fine net. Such a net ought, I 

 think, at all events in autumn, to take the smallest fishes which are then 

 on the bottom, unless perhaps such unusually elongated forms as 

 Lumpenus and pipe-fishes, and young gobies. 



The ideal aimed at is the collection of complete series of fishes with 

 all sizes duly represented, and from this point of view it is worth con- 

 sidering carefully this and other methods employed. 



In many cases the very young fishes are not to be found living on the 

 bottom, and they cannot therefore be taken with a bottom-net, however 

 tine the meshes may be. Such, for example, is notably the case with 

 haddock and whiting long after the post-larval stage, and even 

 certain post-larval flat-fishes, before transformation is completed, may 

 reach a considerable size (up to as much as 40 mm.) while still pelagic. 

 It is evident, therefore, that in order to obtain the early stages in due 

 proportion the method of collection must not be limited to the use of 

 bottom apparatus, but ought to include the use of adequate apparatus for 

 pelagic fishing in the layers of water between the bottom and the surface. 

 It is in this region that the methods of collection have been hitherto 

 least effective. The tow-nets ordinarily employed, with a ring a yard 

 or so in diameter, capture very few post-larval fishes compared with the 

 numbers that we know must be present in the water, and they only occa- 

 sionally take a specimen a little larger. The larger net introduced by 

 Professor M'Intosh for mid-water work, made of mosquito netting, has 

 been more successful, but its comparative delicacy requires that it should 

 be towed slowly, which has the disadvantage of enabling the larger 

 and more active young fishes to escape from it ; and experience on the 

 Garland shows that it is very liable to be ruptured unless in calm weather. 

 It is, moreover, not large enough to be effective for the purpose referred 

 to. On board trawlers I made use of tow-nets composed of the cotton 

 netting described above (B), with meshes about 6 mm. square, and with 

 a ring 5 feet in diameter, without much success, and I have used lately a 

 ring 10 feet in diameter with a similar net and lined inside with 

 mosquito netting. It has been used in winter only, when the mid- 

 water is comparatively barren, but the net itself stood the strain well, 

 and was uninjured. Ring-nets still larger were made use of by Dr. 

 Hjort in the expedition of the Michael Sars,*, and with good 

 success, young cod, haddock, and aaithe being taken in them in the 

 summer. Some of these nets were no less than 21 feet in diameter, 

 and formed of shrimp-netting, and others were 8 feet in diameter. 

 Judging from my experience with the 10-feet ring, nets of such large 

 dimensions must prove troublesome to work when the sea is at all 



*Aarsberetning vedkommende Norges Fiskerier for 1900, 4de Hefte 1900, p. 253. 



