330 Part III. — Tu-entidh Annual Bepoi't 



proportion of round fish when disturbed will rise clear above the wings 

 and escape the mouth of the bag which follows. 



For bottom-fishing, then, the best apparatus appears to be an otter-trawl 

 of the kind above described, and a modification of it seems to be also 

 the best for pelagic fishing. The bottom-net itself — the otter-trawl — 

 when used for pelagic work has one disadvantage that, owing to the retro- 

 cession of the belly of the net and the traction wlien moving through the 

 water, the lower part of the net will be far behind the upper part attached 

 to the headline, and as a fish disturbed at the surface darts downwards, 

 many will escape capture in this way. An ideal pelagic otter-trawl would 

 be one with two pairs of boards, one at each corner, and the mouth of the 

 net in a vertical plane or even sloping backwards and upwards ; but the 

 practical men who make and work the otters are of opinion that four 

 boards would be very diflicult to manipulate. Instead of this, one which 

 is now being made, having a (vertical) mouth of 110 feet in circum- 

 ference, will have the two boards placed at the upper part of the net with 

 a headline of about 25 feet and a detachable bar of similar length affixed 

 to the lower part instead of a ground-rope, the sides of the net being about 

 five fathoms deep. 



Other methods of fishing, such as drift-nets or floating lines, while useful 

 for certain purposes, cannot replace the bag-net method, because they are 

 selective in regard to the sizes, and to some extent the kinds, of fishes 

 they take. 



The Treatment of the Fishes. 



In dealing with the fishes caught, experience has shown the necessity 

 of using a method of measurement as minute and accurate as possible, and 

 as rapid as is consistent with accuracy. Measurements to fractions of an 

 inch, or to centimetres only, are unsatisfactory, more particularly in all 

 except the earlier series. The most important object is to determine as 

 precisely as possible the limits of the various series or generations — the 

 points of division between the groups — and this in many cases is suffi- 

 ciently difficult from natural causes, referred to below, without the addition 

 of difficulties from inexact measurement. In order to do this it is often 

 necessary to adopt a 2-millimetre or 3- millimetre grouping. The fish are 

 therefore measured to millimetres, and although with large or medium- 

 sized fishes the measurements are no doubt in many cases not exact to the 

 particular millimetre taken, the constant endeavour to reach this standard 

 obviously makes the general accuracy greater than if a larger unit were 

 adopted as the standard. 



The method in practice is as follows. A brass measure, one-and-a- 

 quarter inches in breadth, a metre long and divided into millimetres, half- 

 centimetres, and centimetres, the latter being boldly figured, is screwed 

 flush into a groove on the top of a table specially constructed for the 

 work (fig. 2). This table is made of hard wood, is solid and heavy, 

 with outwardly-curving legs, in order to give it greater stability on deck ; 

 it is sixteen inches broad, forty -six inches long, and twenty-six inches 

 high. One end, as shown in the figure, is enclosed by boards to form a 

 receptacle in which the fishes are placed (it might well be larger). Against 

 the end of the millimetre measure, which is fiush with the opposite edge 

 of the table is placed a " nose-piece," so called because the snout of the 

 fish being measured is placed against it. The table also contains a con- 

 venient drawer for instruments, &c., and a sliding rod to which is attached 

 an acetylene lamp that can be fixed at any angle to illuminate any part 

 of the scale at night according to the size of the fishes being dealt with it. 

 The fishes which have been poured on the table at one end are rapidly 



