ANNUAL RINGS IN SCALES. 483 
I give below the average temperature for each month of the hot and 
cold tanks, together with that of the sea water in the tanks at Plymouth. 
LONG: 
a 
July August. September. October. November. December. 
Normal. 15-5° 17:3 15-8 14-6 11:3 10-9 
Hot. 18-6 19-7 19-5 19-8 15-9 17 
Cold. 13-9 13-7 14-1 13-3 10-9 10-3 
1916. 
Jan. Feb. Mar. April. May. June. July. Aug. Sept. Oct. 
Normal. 11-8 10:8 9:7 11:2 13 13-2 Lo iiPigayogs IS) 14-6 
Hot. 16:9) 172) N7-7y ATO Oe LT-9 Ne US 19-6 19:5 16:6 
Cold. 103 95 86 98 11-6 124 13:6 13:7 14 13-6 
It will be seen that as regards the hot tank the temperature throughout 
was practically as high or higher than the highest temperature recorded 
for the normal sea water. Also the temperature was fairly constant from 
month to month. 
As I have already indicated the cold tank records are not as satis- 
factory : in every month the temperature was below that of the normal, 
but in some cases it was above the winter temperature of the normal 
sea water. It was, however, in all cases far below the corresponding 
one for the hot tank. 
99 
‘ABUNDANT’ AND ‘‘ SCANTY ” TANKS. 
These two tanks were the ones in which feeding experiments were con- 
ducted and the temperature was undisturbed. In the abundant tank 
the fish were fed excessively, that is to say, twice a day they were given 
as much food as they would take. The food was varied, sometimes 
squid, at other times worms, etc. 
The fish in the scanty tanks were by no means so well treated : they 
were given very little at a time and never more than once a day : it was 
common to feed every other day. 
In Janvary control fish were started whose scale growth was investi- 
gated in exactly the same way as the others. 
These fish were kept in an aquarium tank together with other species 
and were treated in exactly the same way, being fed once a day neither 
abundantly nor scantily. 
RESULTS OF THE EXPERIMENTS. 
In Tables II, III, IV, V, VI (p. 494), I have given the details of the 
fish experimented upon: the ages as computed from both otolith and 
scale are found in column two ; and it is worth noting here, that the age 
was first determined from the scales and then confirmed by the otolith at 
