82 PKELIMINARY EXPERlMJiNTS ON 



the inner surfaces of which were covered with a layer of hard paraffin 

 wax to prevent the entrance of impurities from the corks in the inverted 

 position in which they were employed. On the fourth day of the 

 experiment (August 20th) three larva in G had completely disappeared, 

 and one had again disappeared three days later (August 23rd) without 

 leaving any trace — a most unusual occurrence. On the 26th, when 

 siphoning off the sediment from the bottom of this jar, I saw a larva 

 sucked into a gap beneath the layer of paraffin covering the cork at the 

 bottom, and was only able to release it with difficulty. The cork had 

 clearly not been driven in quite flush with the outer neck of the jar 

 before the experiment, and had thus been forced inwards to a slight 

 extent after being placed upright in the inverted position, thus detaching 

 the paraffin layer from the inner face of the glass. The crevice so 

 formed was just large enough to allow a larva to enter. This jar was 

 the only one in which such a detachment of the paraffin had taken 

 place, and the defect accounts in all probability for the frequent deaths 

 in G as compared with Y in the early stages, the larvte having got in, 

 and either injured themselves or failed to emerge. After the defect 

 was remedied no further deaths occurred in the jar except one on the 

 following day, which probably resulted from the accident which I had 

 observed. 



Thus exact comparison between the two experiments is scarcely 

 possible, though the healthy development of the larvse in F fortunately 

 renders such comparison unnecessary. 



Ten newly hatched larvre from the same batch of eggs were placed 

 in each of the two jars. They were fed with plankton from the 

 beginning, and half of the water, or even more, according to circum- 

 stances, was changed daily in each jar. A summary of the results is 

 shown in the table (p. 76). 



The advantage of giving the larvse a more plentiful supply of water 

 and food from the beginning was clearly shown by the trifling death- 

 rate at the beginning of the experiment, and by the healthy and 

 vigorous condition of the larvae. In F one corpse was removed on the 

 second day of the experiment, a death which was almost certainly of an 

 accidental character, but from this day no deaths occurred until the end 

 of a fortnight. The larvai were active and healthy during the whole of 

 this period until the last few days, when two, which were not growing 

 so rapidly as the others, showed signs of inactivity and died during the 

 following week. Another was lost on the seventeenth day (September 

 3rd) by being drawn up the siphon during the cleaning of the jar. 

 From this date I was unable to watch the experiments from day to day, 

 owing to absences caused by work at sea and my attendance at the 

 Dover meeting of the British Association. As far as possible, however. 



