i89a. 



THE EVOLUTION OF FLAT-FISHES. 197 



symmetrical. When the lower eye has begun to move towards the 

 edge of the head, the body is held in a slanting position, and 

 throughout the metamorphosis the line joining the two eyes is 

 always kept horizontal, so that when the lower eye is on the edge 

 of the head the body is at an angle of about 45° or more with the 

 vertical. This is particularly well seen in the young turbot, which 

 possesses a swimming bladder, and never seeks the bottom at all 

 until the metamorphosis is almost complete. Giard quotes Mcintosh 

 as witness that a young turbot, 9 millimetres long, with its eye on 

 the edge of the head, swam in the vertical position. This is simply 

 an error of observation. 



We have, however, no satisfactory evidence at present that the 

 monstrous specimens, whose metamorphosis is permanently arrested, 

 swim about in whatever position any more frequently or rest on the 

 bottom less constantly than their norrnal brethren. I have now 

 under observation a living double specimen of the plaice ; its eye is 

 on the edge of the head, the dorsal fin terminates behind the eye, and 

 the posterior three-fourths of the lower side are coloured like the 

 upper, the anterior fourth being white. This specimen, instead of 

 showing a tendency to continue swimming in the water, cannot even 

 be induced to leave the bottom long enough to enable me to see 

 whether it holds itself perfectly horizontal or not. I have never seen 

 it leave the bottom of its own accord ; it lies always buried in the sand 

 up to its eyes, and, when disturbed, makes violent struggles to bury 

 itself again. I have had this specimen for more than six months. 



In spite of all this, Giard not only assumes that double speci- 

 mens swim vertically, but actually maintains that the existence and 

 assumed habits of these double monstrosities is better evidence of 

 the dependence of pigmentation upon light than the result of my 

 experiment. " Dans ce cas comme dans beaucoup d'autres, la nature 

 realise pour celui qui salt la comprendre, une experience bien plus 

 parfaite que celles que nous pouvons imaginer dans nos laboratoires." 

 For Giard goes on to remark that it was only at the price of a 

 thousand difficulties that I succeeded in showing, in an incomplete 

 fashion, that pigment appeared on the lower side of the flounder 

 when light was directed upon it, and that the experiment is not 

 absolutely conclusive. He detects a fallacy in my evidence, and this 

 part of his short paper is still more remarkable than the rest. He 

 points out that, of all the flat-fishes, it is in the flounder that 

 reversed specimens are most common, that is, specimens in which 

 the white side is the right, and the eyes are on the left, instead 

 of vice vevsd, as in normal specimens. He argues that the specimens 

 upon which I experimented may have numbered reversed in- 

 dividuals among their ancestors, and the pigmentation adduced 

 in evidence by me may have been due to inheritance from one 

 of these ancestors which was pigmented on the left side. The 

 probability is the greater in that double individuals are not rare 



