RETROSPECT AND CONCLUSIONS 153 



fication, there is a great similarity of brain pattern, but the relative 

 size of the somatic-sensory and facial or vago-facial lobes varies 

 as the diet passes from a shell-tish diet, through a mixed diet, to 

 a purely fish diet. 



A division of this group may be made by separating the night 

 feeders, the ling and burbot, from haddock, cod, and so on, as in 

 these fish the olfactory lobes become prominent, and the optic lobes 

 less so in accordance with the jiredominance of smell in their search 

 for food. The third cUvision includes two species of an interesting 

 genus physics, or the forkbeards, phycis from the shallow waters of 

 the bays of Madeira, and phycis blemiioides landed at Milford 

 Haven. 



The brain of both fish differs from the usual gadoid type in 

 the largeness of the primitive end-brain and the smallness of the 

 optic lobes ; the former has also a characteristic division. The 

 cerebellum is narrow and very elongated, and the medulla small, 

 with fairly large somatic-sensory lobes. Tliis bottom feeding 

 animal A^liich inhabits rocky ground has its ventral fin on either side 

 converted into a long bifid feeler, and it also has a barbel. Smell 

 and touch wotdd appear to be its chief hunting equipment. Phycis 

 is known as the " schellfisch " by the Germans, and so also is the 

 haddock and phycis blemiioides is also known as the rock haddock ; 

 this is no doubt due to the similarity of diet, but the habitat seems 

 to have given rise to a different adaptation in the forkbeards, as 

 the true haddock, hunts through its taste-buds, and phycis by 

 olfaction and touch, probably as a result of their rocky habitat ; 

 therefore diet has been over-ridden by habitat in determining the 

 method of hunting and brain pattern. 



This seems a favourable opportunity to mention other variations 

 in the primitive end-brain of gadoids ; tliis lobe is always closely 

 connected with the olfactory lobe and varies in size with the im- 

 portance of smell in the hmiting equipment ; the ling and bm-bot 

 emphasise this fact ; when we compare their brains with other 

 gadoids such as the pollack and cod we notice that these lobes in 

 the ling are large and almost hemispherical, approaching the optic 

 lobes in size ; the burbot shows the same relative size of these 

 lobes, as is seen in the ling ; it has been described above that these 

 fish are both night feeders. The primitive end-brain in flat-fish 

 also shows well-marked variations in size. This has been noted 

 in the sole, and the drawing of the brain of this fish shows that it 

 exceeds the optic lobe in diameter and gives rise to olfactory bulbs, 

 wliich tend to overlap. But the members of the groups of flat-fish 

 II and III, in my classification, show throughout a very definite 



