SEA-FISHERIES LABORATORY. Die 
however, it comes out to 0°065644, so that the d, taken as 
trial value (0°02), was evidently too small. The difference 
(0001) between 0°064644 and 0:065644 corresponds to a d 
difference of 0°0003153. Hence the new trial value of 
d 18:— 
0:020000 (originally taken) 
0:00038153 
d = 0:0203153 
This is apphed exactly as d in the previous calcula- 
_ tion and so the d is altered until both sides of the equation 
finally agree. When this is the case the v(s—d,) 1s taken 
and applied in the following :-— 
Volume that would pass throughsopening 
ring if no net attached ; eo: O 
Volume that would pass through opening "~~ v(s—d,). O 
ring when net attached 
@ is the coefficient representing the fraction which the 
area of the cross section of column of water actually fished 
bears to the area of net mouth. For the example taken 
log. u(s). O = 2°51703 
and log. v(s—d,). O= 2°51397 
Hence log. 6 = 0:00306 = 1:00707 
The catch must be multiplied then by 1°00707 to give 
the volume in a column whose cross sectional area is that 
of the net mouth. This reckoning is, however under the 
assumption that the pressure is uniform all over the net. 
. This can be accepted for nets having a large filtering 
area in proportion to the opening, but it is evident from 
a series of catches made by Hensen with the same net 
and gradually increased area of mouth opening, that the 
larger the mouth, the more incorrect does the @ become. 
