15 
.rawling was continued until the time came when it was necessary 
,o return to Blyth to catch the train leaving at 9-20 p.m. During 
ihe day a numher of the fish were examined with regard to their food 
md maturity, and in recent years many of the fish caught were 
•etaiucd after each haul to he marked and returned to the water. 
\t the end of the day's trawling all the marketable fish were counted. 
The small examples were returned after every haul, and as I have 
pointed out before in a living condition. 
On the few occasions that the first haul showed that the ground 
was in an abnormal condition, usually from the presence of great 
quantities of weed, another station was experimented in instead. 
The two most northerly stations were trawled in from early morning 
to late afternoon. 
Marketable Fish. L — Seasonal Variation. 
It is proposed to deal first with the marketable flat fish, that is 
to say, with the fish referred to above which were counted at the end 
of an experiment lasting usually from seven to ten hours. 
In previous reports, bearing in mind the regularity with which 
the experiments were made in relation to fixed trains, the results 
were presented in the form of the total catch for the day in each case. 
For the purpose, however, of instituting a strict comparison between 
the various stations, and with the records of the first haul, which 
will be dealt with next, the results are now expressed in common 
terms, viz : the catch per hour's trawling. 
A summary of the catches of flat fish for the period of 1-4 years 
is given in liable IV., which also shows the mean catch per hour for 
that period. 
The details of all the experiments are given in Table V., viz. : 
the catch per hour for each experiment, the mean catch per hour for 
each station, and the mean catch per hour for the area each year. 
As each station was visited each season once, twice, or at most 
three times, and was trawled in for a period of many hours, it is not 
possible to determine their character by reference to a large number 
of experiments for any one season, or to a part of a season. The 
degree of variability may be estimated, however, in one or two cases 
where the experiments were repeated within a short period. For 
example in 1902, Cambois Bay was experimented in on June 28th 
(8 hours), on July 2nd (9| hours), and on July 23rd (9f hours). 
The nuan catch per hour on these dates was : plaice, 15-13, 9*95, 
;1 »<1 14-05 ; dab. 11-88, 16-48, and 10-81. Again in 1892, the same 
bay was visited on August 11th (10 hours), on September 13th 
