1894.] TAEIATIOK IN FLAT-FISHES. 247 



upon examination found to be an expression of the fact that the 

 ordinal positions of the neural spines crossing the centres of the 

 spots of the dorsal row are, if reckoned from behind forwards, 

 almost exactly the same as those of the haemal spines crossing the 

 centres of the ventral spots. 



The particulars are as follows : — The centre of the most posterior 

 dorsal spot stands almost exactly over the neural spine of the 11th 

 fin-ray, reckoning from behind. The centre of the next spot is 

 upon that of the 26th fin-ray reckoned from behind ; that of the 

 next is on the 42nd. 



Of the ventral spots the centre of the most posterior is on 

 the haemal spine of the 11th fin-ray from behind ; that of the 

 next is on the 24th, and that of the most anterior is on the 

 40th. It will be seen that the numbers in the two rows closely 

 correspond. 



Continuing the dorsal series there is a spot on the neural spine 

 of the 52nd fin-ray, another very minute and faint spot over the 

 63rd. In front of this there are 16 more fin-rays. 



The whole number of fin-rays in the dorsal fin is 79, and in the 

 anal fin 58. 



On detailed examination it may be seen tbat the spots are not 

 wholly shapeless blotches of colour, but that some of them con- 

 sist of several irregular zones of colours. Each of these spots is 

 thus a somewhat indefinite ocellar mark. The spots dd, vc, and 

 vb have each a minute centre of light colour, which is chiefly due 

 to the presence of a whitish scale in the middle of the spot. This 

 light colour is not altogether confioed to the one scale, but spreads 

 a little on to the edges of the adjacent scales. The spot db has 

 two of these small whitish centres. The spot dc alone of the five 

 chief spots has no light centre. Around the centre of each spot 

 are scales of a brown tint bearing specks of very dark pigment. 

 These deeply pigmented scales form a zone about four rows deep 

 in the case of the larger spots. Outside this is an irregular zone 

 of fine pigment-granules giving a neutral tint. Beyond this again 

 there was in the case of spots dd, dc, db, and vc a vague and 

 imperfect band of silvering, forming a border to the proximal limbs 

 of the spots. 



It is thus seen that the colours of the dorsal and ventral bordei*s 

 have, so far as the last three spots are concerned, varied similarly 

 and simultaneously, producing a result that is nearly symmetrical 

 about the horizontal axis. This phenomenon is precisely compar- 

 able with the much more common phenomenon of similar and 

 simultaneous variation of the right and left sides of a bilaterally 

 symmetrical body. It is to be remembered that in many fishes, 

 and especially in Flat-fishes, there is an imperfect relation of bilateral 

 symmetry subsisting between the parts dorsal and ventral to the 

 horizontal median plane. This symmetry is generally manifested 

 both in form and colour, and is an indication that at some time 

 these parts have undergone similar variation. The present ex- 

 ample illustrates the principle that parts, which in the normal are in 



17* 



