SEA-FISHERIES LABORATORY. 148 
that it is undesirable to use the tow-nets over the stern on 
account of the mixing of the layers of water caused by 
the propellers. It is also bad practice to tow the net 
directly from the ship’s side. It is very desirable to use 
the net always on the windward side. There appears to 
be a difference in the plankton taken to windward and 
leeward—that is if the ship is making any leeway. A 
small otter-board of about two feet square is trimmed so 
as to tow well out from the ship’s side, and the tow-nets 
are fastened to this. By this arrangement the net tows 
about five or six feet out from the side, and it is practically 
unaffected by rolling or pitching which, when the net is 
towed from a rope fastened to the side directly, are apt 
to cause the net partially to rise out of the water, and 
vice versa. 
Deep hauls with the ordinary tow-net were formerly 
made simply by attaching the net to the back of the trawl, 
using a short rope so that the net could not roll off the 
latter. But this did not, of course, give a really bottom 
sample. It was useful for securing a rough sample of the 
general plankton contents of the water, and this is, of 
course, all that can be expected from hauls taken by 
means of nets unprovided with conical head pieces. By 
far the most accurate means of explorifg the vertical 
distribution of the plankton is that of taking samples of 
the water at different depths by means of the water-bottle, 
and we have practised this on many occasions. 
Unfortunately the capacity of the Nansen-Pettersson 
bottle 1s too small in general to afford a big catch. It 
would be an advantage if some water-bottle were designed 
for this purpose alone. Kofoid has indeed suggested a 
form of water-bottle for collecting plankton, but the 
apparatus appears to us to be rather costly. 
The vertical nets are worked in the same manner and 
