AT THE PLYMOUTH LABORATOEY. 307 



was from two to three days, and the whole of them were thus trans- 

 ferred to fresh water by November 15th. On November 9th, shortly 

 after the change to fresh water, one of the males died. The testis 

 was found to be nearly mature. The weight of this fish was very 

 nearly 1 lb., and the length 13f inches. During this period one of the 

 fish jumped from the fresh-water tank into a tank of sea water 

 adjoining (cp. p. 305). After thirty-six hours it was put back into 

 the fresh-water tank direct without any ill effects arising. 



November 26th. All of the grilse were examined by Mr. McNicol, 

 who has charge of the Duke of Bedford's Fishery. Apart from the deaths 

 that had been recorded, it was found that five of the fish were missing. 

 It can only be surmised that these had from time to time jumped over 

 the barrier into the large adjoining tank and fallen victims to the 

 turbot, nurse-hounds, and other large fish that occupied it. The 

 number that remained was now thirty-five. Seven females were 

 spawned on this day (November 26th), and the ova fertilized. The 

 remainder were spawned on December 11th. Nearly the whole of 

 these fertilized ova were taken to Endsleigh and there hatched under 

 normal conditions both as regards numerical proportion and the 

 period between fertilization and hatching. They were not kept under 

 observation for long after hatching, and there was apparently no 

 unusual feature arising in regard to size or otherwise in connection 

 with their development. A few ova were retained at the Laboratory 

 and hatched under tap water, but these did not long survive the 

 feeding stage following absorption of the yolk-sac, owing no doubt 

 merely to the unsuitable condition of the water supply with its 

 irresistible tendency to nurture the growth of fungus and other 

 vegetable organisms. 



As regards feeding during this period of spawning, very little food 

 was taken between the last week in October and the two respective 

 dates of spawning, November 26th and December 11th, though feeding 

 did not entirely cease for more than a day or so at a time. Within 

 a few days after spawning, there was a marked change in this respect 

 and by December 22nd the total consumption amounted to about f lb. 

 of liver per day, the weight of the fish at this time averaging about 

 1^ lb. Feeding continued at much the same point till the middle of 

 February. It was then decided to turn the fish, numbering thirty-five, 

 over to sea water, and this was done between February 19th and 

 20th. The change proved to be an unfortunate one, and the salmon at 

 once ceased feeding. Between February 21st and 26th, seven of them 

 died — five males and two females. These were found to still contain 

 a quantity of ripe milt and ova respectively. Between February 25th 



