of Edinburgh, Session 1884-85. 
13 
grew out to meet the epidermis at the sides of the head. But 
Sedgwick’s revelation that the floor of the cerebral vesicle was once 
the surface of the head, enables us to go a step further, and form a 
clear conception of the state of the eyes in the primitive vertebrate 
ancestor, and the transition to the present state. In the worm-like 
ancestor at the sides of and somewhat in front of the mouth were 
a pair of eyes, formed as simple pits in the epidermal nerve area, 
exactly similar to the simple open eye-cups which exist in Patella 
and Haliotis at the present day. When the mouth and nerve area 
were covered by the formation of the neural canal the eyes became 
cerebral, and were influenced by light coming through the trans- 
parent tissues. In this condition the eye has persisted in Ascidians. 
In Vertebrates the body has become opaque, the eye-cups have 
grown out towards the side of the head, where a thickening took 
place in the epidermis to form the crystalline lens, which afterwards 
was pinched off and sunk in the mesoblast. We may regard the 
humours, the cornea, and the lens of the vertebrate eye as parts which 
have retained their primitive transparency (see Plate II. fig. 6). 
We have now made more compact the inversion theory of vertebrate 
morphology. Sedgwick has attempted to go further, and account for 
segmentation gill-clefts, segmental organs, and abdominal pores, by 
deriving them from Coelenterates similar to Zoantharians. His 
theory is very ingenious, and for the most part extremely probable. 
But it agrees much better with the view that the infundibulum 
represents the primitive mouth, and the neurenteric canal the primi- 
tive anus, than with Sedgwick’s own view concerning these points. 
For, according to Sedgwick himself, the planes of segmentation are 
perpendicular to the direction of the elongated blastopore, while if 
the present anus of a fish is the primitive anus, the segments of the 
fish’s tail are parallel to part of the primitive blastopore. And if 
the primitive blastopore extended from the actual mouth to the 
actual anus in such a fish as a cod or blenny, it would occupy 
almost y^-ths longest circumference of the body. Balano- 
glossus, too, seems to give evidence that openings can arise from the 
intestine to the exterior independently both of segmental organs and 
blastopore. There is one other point which Sedgwick neglects, 
namely, that segments in all segmented animals are formed between 
the telson and the last segment. I am not aware that in any Coelen- 
terate the mesenteries are formed according to this law. 
