AGE AND GROWTH OF THE CISCO 
295 
catches of herring in drift nets of three sizes of mesh. 34 They summarized their results 
as follows: 
According to these samples the finer-meshed nets caught the larger fish. The conclusion to be 
drawn is that the size distribution of the drift-net catch is probably not determined to any great 
extent by the selective action of the net. 
Buchanan-Wollaston (1927) included with his discussion of the selective action 
of a trawl net certain observations as to the method of selection in drift nets. He 
called attention to the fact that drift nets do not capture all fish in the same manner. 
He pointed out that certain large individuals are retained by the maxillary barbs; 
some small individuals are held only because of the gill covers; while the majority 
are held between the gill covers and the dorsal fin. The different types of net action 
may lead to discontinuity in the length distribution of the catch. 
Hodgson (1927) presented the analysis of data from the catch of herrings in 
three drift nets of different-size mesh. 35 Although the number of specimens was 
not great his results showed clearly that selective net action can lead to erroneous 
conclusions concerning the size and age composition of a population and also lead 
to inaccurate determinations of average lengths for the various age groups. Later 
the same author (Hodgson, 1933) presented the results of further experiments on 
the selective action of drift nets. These experiments were conducted with particular 
reference to the effect on the herring stock of the size of mesh in the drift nets em- 
ployed in the commercial fishery. On the basis of two separate experiments Hodgson 
concluded: 
that there is a very definite and subtle gradation in the length of the herrings caught by nets of 
even slightly different meshes, and it is also plain that the whole character of a fishery can be changed 
by the use of different nets. 
These investigations were all concerned with a single marine fishery — the drift- 
net fishery for herring. The opinions of the different investigators, however, show 
considerable disagreement. 
Of greater interest with respect to the present study are the data concerning 
the selective action of gill nets as they are fished in fresh water. In the following 
paragraphs certain of the available data with respect to the selective action of gill 
nets are mentioned briefly. The various investigations are treated in general in the 
chronological order of their appearance without any attempt at grouping according 
to variety or kind of fish concerned. 
Jarvi (1920) compared the catches of the kleine Marane ( Coregonus albula ) 
taken by seine with those taken by a series of seven nets having meshes ranging 
from 10 to 15 millimeters (bar measurement). The length compositions of the 
samples from the two sources agreed well. 
Jarnefelt (1921) noticed that his growth curves for several species of fish 
showed irregularities that could be traced to the selective action of the nets that 
he used in obtaining his samples. Concerning these irregularities Jarnefelt stated: 
“* * * dass es ebenso viel Senkungen der Kurve gibt wie Netze verschiedener 
Maschenweite benutzt werden.” However, Segerstr&le (1933) held that the irregu- 
larities observed by Jarnefelt and others in growth curves resulted to a large extent 
from inaccurate methods of calculating growth from scale measurements, and that 
88 The sizes were 30, 31, and 33 meshes to the yard (bar measurement). 
83 The nets used had 35, 38, and 48 meshes to the yard (bar measurement). 
