ix 1'KOTOSTOMA THEOKY 495 



neural rudiments and two notochords of half the usual si/..-, widely 

 separated by the mass of yolk or endoderni. 



Similar al-norma lilies have been observed in Teleostean fishes. 

 E.g. Fig. 220, D shows a Pike-embryo which is normal towards its 

 anterior and posterior ends but interrupted for some distance in the 

 mid-dorsal line by a wide cleft in which the yolk is visible. Again 

 in l-'ig. 220, C a similar cleft is seen to traverse the whole length of 

 a Trout-embryo from the hind-brain region tailwards. 



An important feature of such abnormal embryos of fish and am- 

 phibians is that they frequently proceed with their development, the 

 lips of the fissure closing up and the two sets of half-organs being 

 brought together in the mesial plane, undergoing complete fusion 

 and the individual becoming in fact entirely normal. The import- 

 ance of this return to the normal on the part of such split embryos is 

 that it indicates that the departure from the normal during the split 

 condition is far less fundamental than would appear at first sight. 



Here then we have to do with two very remarkable phenomena. 

 Firstly there is the abnormality itself the fact that the dorsal region 

 of the body is for a time in the form of two distinct halves. Abnor- 

 malities of such a definite type as this usually have a definite evolu- 

 tionary or other meaning and it is necessary to search for such a 

 meaning in this particular case. Secondly, there is the fact that an 

 embryo almost completely bisected in this way is frequently able to 

 right itself and become perfectly normal. This again suggests the 

 question whether this power of righting itself has not some special 

 evolutionary meaning. 



III. In the higher meroblastic Vertebrates we have seen that 

 there exists along the middle line of what corresponds with the 

 archenteric roof of Amphioxus (i.e. the region which becomes con- 

 verted into the dorsal part of , the body, including notochord and 

 medullary plate) the structure known as the primitive streak. We 

 have also seen that in the lowest Vertebrates possessing it, this 

 primitive streak represents the line of fusion of the gastrular lips, 

 and that we are therefore justified in attaching the same significance 

 to the primitive streak in those higher forms in which the actual 

 process of fusion can no longer be observed. That this interpretation 

 is correct is indicated by the occasional occurrence of openings in the 

 line of the primitive streak communicating ventrally with the enteron 

 and dnrsally with the outer surface of the medullary plate, or its 

 derivative the floor of the neural tube (pp. 51, 53). Such neurenteric 

 communications are readily explicable by the view that they repre- 

 sent simply parts of the line of fusion of the gastrular lips where 

 the actual fusion has not been completed. Here again we have a 

 phenomenon which demands explanation the occurrence of what 

 seems to be the vestige of a slit-like gastrular mouth along the mid- 

 dorsal line. 



IV. We have another remarkable body of facts associated 

 with the fate of the blastopore or remnant of the gastrular mouth in 



