c^ the Fisher II Board for Scotland. 161 



I have tabulated the temperature observations in the accompanying 

 tables for each of the tanks, and for the supply as it came to the 

 apartment. Tanks No8. 1 and 2 were not supplied with hot water; Tank 

 No. 3 got a partial supply, and Tank No. 4 the largest supply, and it is 

 this tank which was looked to to give the best results. 



The temperature observations are tabulated in ten-day periods for the 

 155 days over which the experiment extended, each showing the 

 maximum and minimum temperature recorded, the mean for the ten 

 days ; and for Tanks 3 and 4 the mean of the maxima and minima have 

 also been calculated. 



It will be seen from the tables that the mean temperature in No. 1 

 varied from 3-9 C. to 10-3 C. during the 155 days, the mean for the 

 whole period being 6*5 C. (43 '7 F.) ; the extremes of temperature 

 observed at any time were 2*8 and 10"6. The mean temperature of No. 

 2 varied from 4-3 C. to 10"8 C, the mean for the period being 7"0 C. 

 (44-6 F.) ; the slightly higher temperature was owing to this tank being 

 nearer the stove that No. 1. In No. 3 it was desired to maintain a 

 temperature intermediate between that of No. 4 and the other tanks, but 

 greater difficulty was found in this case in adjusting the supplies of hot 

 and cold water. The ten-day means ranged from 6 C. to 11 '9 C, the 

 mean for the period being 8-6 C. (47 "5 F.). The extreme individual 

 temperatures were 3 "9 and 15 "6, a difierence of 11*7 C, but these 

 variations were of short duration. In No. 4 the range of the ten-day 

 means was lO'S to 13-6 C, the range of the mean maxima was from 17'2 

 to 12*6, and of the minima from B'O to 11"8; the extreme temperatures 

 recorded were 6-8 C. (44-2 F.) and 22-8 C. (73-0 F.), a difference of 16 0. 

 or 28"8 F. The low temperature as a rule occurred when the stove or 

 some part of the apparatus required to be overhauled, and the high ones 

 for a short period, when the fire had been too strong ; they sometimes 

 occurred during the night. The mean temperature for this tank for the 

 whole period was 12*5 C. (54*5 F.), which approximates to the mean 

 bottom temperature in the sea off the East Coast in July, August, and 

 September ; in depths of from ten to fifteen fathoms the mean temper- 

 ature for these months is about 52*9, and a little further out, in thirty 

 fathoms, it is 50' 7 F. 



The fishes used in the experiments were young whitings, codlings, and 

 haddocks, a few dabs, a plaice, a small starry ray, and an armed bull- 

 head. They were procured in the small-meshed net used around the 

 otter-trawl in the investigations made on board trawlers, and were first 

 kept for a few weeks after being brought to the laboratory before they 

 were placed in the experimental tanks. Each fish was measured, but 

 not weighed, nor was its volume determined ; trial showed that the risks 

 might be too great. 



"With regard to the general conditions and behaviour of the fishes a 

 little may be said. They were fed daily, or several times a day, and in 

 all cases they got as much food as they were willing or able to take, but 

 they were not fed during the night. Their food consisted almost 

 entirely of the ordinary edible mussel, chopped up, varied occasionally 

 ■with a few limpets, and still more rarely with live shrimps ; on one or 

 two occasions they got fragments of herrings or parts of the roe or milt, 

 and sometimes the mussels were not removed from their shells, but were 

 broken up and crushed. An attempt was at first made to weigh the 

 quantity of food given to them daily, but the conditions of the experi- 

 ment showed that this might be misleading and it was discontinued. 



All the fishes did not by any means thrive to a like extent. The 

 haddocks, in particular, proved to be exceptionally delicate as compared 



L 



