RESEARCHES ON RACES OF HERRINGS. 83 
this way the insertions of the keeled scales could be made out and their 
total number established with certainty. Practically no difficulty was 
experienced in counting the rays in the pectoral fin (12). It was noticed, 
however, that in fins with a large number of rays the increase in the 
number appeared to be accounted for by additions of small rays near 
the posterior border. No attempt was made to analyse this fin as in the 
case of the dorsal and anal fins, but it is possible that useful information 
might be obtained by attempting such an analysis. 
The weight of the fish (15) was taken separately and to the nearest 
gram. 
In recording the condition of the gonad (16) it was found necessary to 
use combinations of the numerals representing different stages, which 
require explaining. 
In the records occur such combinations as [V-VI and VI-IV. A record 
such as [V—VI is put down to represent fish in which the gonad appeared 
to be about ripe, although it did not fill the body cavity entirely. These 
records, however, refer mostly to males, in which the approach to and 
incidence of ripeness of the gonad are not easy to differentiate. In the 
case of records such as VI-III or VI-IV, these indicate that the gonad is 
definitely ripe, but has become reduced by spawning (or compression in 
some cases) to the size in the stage indicated by the second numeral ; 
thus VI-LV indicates gonad reduced to the size of half the volume of the 
abdominal cavity. The numeral VII was reserved for fish which were 
spent or practically spent. By distinguishing spawning fish in this way 
it is possible to correlate to some extent the weight with the size of the 
fish. 
With regard to Characters | to 9 all measurements were taken with the 
instrument suppled by the Board of Agriculture and Fisheries. An 
attempt was made to measure Character | to the nearest -2 of a millimetre 
im Samples III and IV. 
Character 2 was measured to the nearest -5 mm. in Sample III and to 
‘2mm. in Sample IV. 
Characters 3 to 9 were measured to the nearest -5 mm. in both 
Samples IIT and IV. 
DEFINITION OF CHARACTER 7. 
Characters 1 to 9 are those recommended by the Board except No. 7. 
Character 7 is defined in the Scheme as “ from the end-board, ete. . . . to 
the posterior edge of the hindmost scale.” In preliminary investigations, 
however, it was found that posterior scales were either rubbed loose or 
missing in about one-third of the specimens examined. It was therefore 
