28 



of 15°. The top and sides of the vessel containing the fish were covered 

 with opaque material, and through the cover were passed a jet deliver- 

 ing water and an outflow pipe connected with an overflowing bottle a 

 little distance off, so that a constant circulation of sea water was main- 

 tained in the vessel while the level of the water in it remained con- 

 stant. The whole apparatus was placed in front of a south window 

 from which the light fell on the mirror and was reflected vertically 

 upwards on to the bottom of the vessel containing the fish: as the fish 

 were usually resting on this bottom their lower sides were illuminated 

 while their upper sides were in the dark. 



At the same time I kept a large number of the same young floun- 

 ders living under ordinary conditions in table-tanks at the bottom of 

 which was a layer of fine sand. 



I fed these young flounders first with minute Crustacea sifted out 

 from weeds gathered on the shore, and afterwards with minced worm, 

 and they all throve well and grew rapidly. 



On June 21st I took out one of the specimens from the mirror- 

 apparatus and examined it. It was 2-7 cm in length. Another specimen 

 taken from an ordinary tank for comparison was 2-6 cm long. The 

 difference between the lower sides of these two was as follows: In the 

 mirror-specimen there was an opaque white layer all over the wall of 

 the abdominal cavity, the rest of the skin being translucent. In the 

 normal specimen this coating was confined to the edges of the same 

 area. There were a few scattered black chromatophores on the lower 

 side of the head in each specimen, but rather more in the mirror spe- 

 cimen than in the other. 



It is evident that these differences are not very important, and I 

 think it is reasonable to conclude that at this time, one month and a 

 half after the commencement of the experiment, the lower sides of the 

 mirror-specimens had become, by inherited tendency, as destitute of 

 pigment as those of the specimens under the ordinary conditions. 



liut two months afterwards namely on August 27th all the flounders 

 in the mirror apparatus died. The cause of death was this. After my 

 return from Norway on August 13th I noticed that the fish in the 

 apparatus very frequently clung to the sides of the vessel instead of 

 lying on the bottom, and as the sides were darkened, while they were 

 in this position their upper sides only were exposed to the light from 

 the mirror. In order to prevent this I introduced a horizontal partition 

 of network so as to keep the fish on the bottom of the vessel ; but the 

 netting soon got obstructed with remains of the food, and the water 

 below the partition was thus cut off from the circulation so that the 

 fish were asphyxiated. 



