88 PKELIMINARY EXPERIMENTS ON 



from about 16 mm. to 22 mm. in length. The actual dimensions a 

 fortnight later are given below. 



Up to this time (the end of October) the fry had been regularly fed 

 with plankton, but at the end of this period the available supplies of 

 large Copepods became much reduced, partly from the impoverishment 

 of the plankton which normally takes place in autumn, and partly from 

 bad weather in November, which interfered with the work of collection. 

 The fry would no longer attack the smaller Copepods {Oithona, Para- 

 calanus, Acartia, etc.), and Calanus was not obtainable. No Crabs, 

 Prawns, or Shrimps could be found in berry, except isolated specimens 

 carrying new-laid eggs, which the fish refused to touch. I therefore 

 tried feeding them with bits of chopped Prawn. They would occasion- 

 ally seize morsels as they fell through the water, but not after the 

 fragments had settled on the bottom. On the 6th of November one of 

 the largest Blennies died, almost certainly of starvation, owing to the 

 difficulty of providing suitable food. On the 9th I introduced a 

 number of small Crustacea (Mysidas, Cumacea, and Amphipods) 

 obtained by working a tow-net immediately above the sandy bottom 

 in Cawsand Bay. A more successful diet was provided on the follow- 

 ing day by working a cheese-cloth net over the Zostera bed in the same 

 localitv, which brought in a number of small Terebellid and Nereid 

 worms with a few Crustacea. The results were excellent, for the 

 wriggling movements of the worms after they were thrown into the 

 jar at once attracted the attention of the fishes, which snapped them up 

 with avidity. The Nereids appeared to prick the fishes and to be 

 relatively unpalatable, but the Terebellids were eaten with great 

 satisfaction. The largest of the surviving Blennies, which had been 

 ailing during the past week, and had refused all food, now began to 

 recover, and shared the new food with the others. I saw it capture 

 quite a large Amphipod, which, after a little difficulty, it succeeded in 

 swallowing. Only one further incident demands notice. On the 8th of 

 November, having observed that for several days past a piece of Ulva 

 frequently had one of the fishes underneath it, I introduced a number 

 of small empty Gastropod shells, chiefly Trochus, into the jar. One of 

 the little Blennies at once tucked himself inside one, resting on his 

 pelvic fins, with his head outwards near the orifice, in all respects like 

 the adult fish. A little later there were two fish inhabiting shells. 

 The adoption of this habit at so early a period was not only a sign that 

 the fry had completely departed from the pelagic habits of the larval 

 stages, but is interesting from a general point of view. It shows that 

 the conchicolous habit of the adult fish is not a mere nesting device for 

 the protection of the eggs, but is adopted at probably all stages of 

 growth for concealment and protection. 



