66 /. H. Powers 



quently from chance head-on colHsions, resulting from the 

 Chorophilus' own movements. 



Last of all, before leaving this type of variation, I will add that 

 the modifications in external characters of which I have spoken 

 and which are shown in the plates are no more important than 

 are other and inner modifications. Reserving for a fuller pre- 

 sentation the details of most modifications, I will state a few of 

 them simply. That muscles are elongated and otherwise modi- 

 fied is obvious. Most of them are not proportionately developed 

 in breadth. The lower jaw becomes disproportionately heavy. 

 The gill arches are not only greatly elongated but the gill rakers 

 become separated by nearly twice the interval which separates 

 those of the ordinary larva. Even the brain, easily visible 

 through the soft palate, which is especially transparent in the 

 cannibals, is modified. Instead of growing with the growth of 

 the head, it is noticeably smaller, not only relatively but abso- 

 lutely, and of a less compact and more piscine type. Its position, 

 too, with regard to certain portions of the skull, is surprisingly 

 changed, the change being apparently due to the development of 

 the anterior part of the skull itself. Thus in a normal larva of 

 about lo cm. the brain reaches to, within 5 mm. of the snout, 

 while in a cannibal of the same length the distance may be in- 

 creased to nearly or quite 8 mm. And if, in place of this absolute 

 displacement we observe the displacement relative to certain 

 other organs the modification becomes much more conspicuous, 

 e. g. the cerebral hemispheres in the normal larva extend well 

 forward between the lateral combs of the palatine teeth. If the 

 head were sectioned just posterior to the palatine teeth, the line 

 of section would pass almost or exactly between the bases of the 

 cerebral hemispheres and the optic lobes. On the other hand, in 

 a cannibal the brain does not extend between the palatine teeth 

 at all, and a section removing the anterior portion of the head 

 might include all of the palatine teeth and something more with- 

 out touching the brain. This relative displacement is, as I have 

 indicated, due to a modification of all the parts and not to a mod- 

 ification and displacement of any one. 



262 



