4 Fhherii Jhnrd for SrofJavrJ. 



was not attached to the gravel. Some eggs had been killed. A fair 

 amount of ova was still adhering to the stones. The}- liatched out 

 well, as did some of the eggs whicli were found loose among the 

 gravel. The spawn was divided up. Part of both lots was cooled, 

 while the remainder was kept in uncooled water. The glass plates, 

 both cooled and uncooled, were made to revolve. The spawn on 

 gravel, both cooled and uncooled, was put into trays, which were 

 moved vertically. The revolution of the plates, and the raising and 

 depressing of the trays, was intermittent, not continuous. The trays 

 were made of wood, and they had bottoms sparred with glass tubes, 

 and in one case covered with old galvanized wire nettintj'. 



The uncooled glass plates were put into the apparatus used in the 

 1910 experiment.* It consisted of a galvanized iron box, which 

 revolved inside a galvanized iron tank. 



The apparatus used for the cooled spawn was of wood, most of 

 which was tarred. The box containing the plates was 21 inches 

 square and 18 inches deep (A., fig. 2). A wooden frame, which 

 carried a plate on each of its four external faces, revolved on iron 

 axles. .Each plate was in succession brought to meet the inflowing 

 current of water. 



In the case of the uncooled gravel spawn the trays were raised a 

 little and then depressed. The cooled gravel spawn was arranged in 

 trays that travelled roand on a revolving frame (B., fig. 2). The 

 trays were 9 inches by 7 inches by 2 inches deep. 



The water was filtered throngh a barrel filled with sand, both for 

 cooled and uncooled ova. 



For cooling the water a series of galvanized iron pipes (|-iuch wide 

 diameter), covered with ice, was employed. 



Sea-water had been running through the pipes and the apparatus 

 for three months before the ova were introduced. 



The spawn was put into running water b}' about 10 p.m. on the 

 28th February. 



The temperature of the uncooled and cooled water, and the 

 C[uantity of water supplied to several of the lots of eggs, are shown in 

 the table, p. 5. The temperature in °C. was read almost hourly 

 during the day and night. Tlie maximum and minimum thermo- 

 meters were read once, sometimes twice, a ^ay. The temperatures 

 given in the maximum and minimum columns refer to parts of two 

 days. The quantitj- of water is indicated by the time required to fill 

 a one-pint measure (=*6 litre). 



An estimate was made of the number of eggs on two plates, which 

 were not too crowded with ova. They contained approximately 2600 

 and 2800 eggs respectively. 



Rate of Movement of the Apparatus. 



The cooled glass plates revolved once in from 1 to 2 minutes. The 

 cooled gravel in traj's made a revolution in from 5 to 10^ minutes. 

 The plates in the tin revolved once in from 1^ to 2 minutes. The 

 uncooled gravel spawn in trays was raised and depressed once in 5i 

 to 7 1 minutes. 



* " Experiment in x'etarding the Development of tlie Eggs of the Herring.'' 

 Twenftj-eighth Annual Report of the Fi-fhfnj Board for Scotland, Pai't III., 1911. 



