RESEARCHES ON RACES OF HERRINGS. 75 
were damaged with respect to one or more of the characters required, and 
were rejected ; apart from these damaged fish there was no other selection 
effected. 
Both larger samples were samples from a large haul of fish and were 
taken at random from the catch. 
The method of work in detail was as follows :— 
Ist series of operations. 
The fish were first weighed singly and a sample of the scales taken 
from the region under the pectoral fin and put in a previously numbered 
envelope. A light metal label attached to a small safety pin was then 
stuck into the fish, which was passed on to be measured, the weight of the 
fish in the meantime being called out to the recorder. The measurements 
1 to 9 were then taken—being called out and recorded successively. The 
sex and condition of the gonad were next determined and recorded and 
the fish finally labelled—by pinning the label to the skull through the 
orbit—and put into an ice-chest tray. Four persons were concerned in 
this operation—one to weigh and take scales, one to record, one to hand 
the fish on and assist mechanically with the measuring, and one to measure 
and take condition of gonad and sex. It was found that weighing and 
taking scales could be done on the whole rather more quickly than taking 
measurements and sex. In this way from about 34 to 44 fish could be 
examined in one hour’s continuous work. The time within which a given 
number of fish were examined was noted and is given in the account of 
the examination of the different samples. 
2nd series of operations - Counting. 
In the large samples the counting of the keeled scales and fin rays 
(Characters 11 to 14) began on the third day of the investigation and was 
finished on the fourth. When four workers were available each counted 
the same fin—or the scales—in all the fish, and handed each fish on to 
his or her neighbour in turn. In the early stages of the work each worker 
called out the count to be recorded, and later each worker kept a record 
to check the count called out, but mm the whole of the second sample, 
each worker simply recorded his or her own work. When the Characters 
11 to 14 had been recorded the alimentary canal was taken out and 
labelled with the number of the fish, and preserved for the future examina- 
tion of the pyloric ceca. 
3rd stage : Preparing skeletons. 
It was found better to place the fish in cold water, to bring the water 
to the boil, and allow to boil only two minutes than to boil for ten minutes. 
