ATTEMPT TO DEDUCE VERTEBRATE EYES FROM THE SKIN. 3865 
significance to find that, in these same animals, not only does 
the nervous system arise as a solid strand of ectoderm, but the 
eye and the optic nerve arise “‘ quite high up on the side of 
the brain,” ? and only sink down later into the ventral position 
(with respect to the medullary axis) typical of the higher 
Vertebrates. It seems to me that these facts belong to one 
another, and may be claimed as primitive features retained 
by these lowly Vertebrates.? 
It is, I am aware, usual to regard this development of the 
medullary axis as a solid strand as secondary, but the argu- 
ments are not conclusive. For the reasons above given, I am 
inclined to believe that the solid medullary axis in these cyclo- 
stomes and bony fishes is a direct inheritance from their Inver- 
tebrate ancestors ; and that, conversely, the widely open medul- 
lary groove is a secondary specialisation for the purpose of 
supplying a still larger quantity of material as the nervous 
axis of the Vertebrates became more and more pronounced.® 
According, then, to our interpretation of the facts, the 
1 Hoffmann, ‘ Arch. mik. Anat.,’ xxiii, 1883, p. 45. 
* Evidence of this secondary connection of the optic invagination with the 
brain has, further, been recently deduced from a study of the developing 
brain itself. Waters, from a study of the primitive segmentation of the brain 
(‘ Quart. Journ. Micr. Sci.,’ vol. xxiii, 1892, p. 457), arrives at the conclusion 
that the optic nerve was once serial with the other segmental nerves. The 
fact that the larger proportion of the nerve fibrils develop from the retina 
towards the brain [Assheton, ‘Quart. Journ. Micr. Sci.,’ 34 (1892), p. 85; 
Robinson, ‘ Journ. Anat, Phys.,’ 30, 1896, p. 319] may perhaps represent 
the enormous secondary multiplication of the retinal sensory cells. 
3 The presence of a medullary plate in forms which appear to be still 
lower in organisation than the Cyclostomes is perhaps a difficulty, but the 
very diversity of their specialisations, Urochorda, Cephalochorda, Hemichorda 
(see Lankester’s article “ Vertebrata,” ‘Encyc. Brit.’), suggests that they have 
wandered off in different directions—all, however, towards degeneration— 
from some more typical Vertebrate form. The lowness of their organisation 
may perhaps be paralleled by that of the Copepods, which are apparently 
simpler than the form Apus, which has the best claim to rank as primitive 
among the Crustacea. I would suggest a similar explanation for both cases, 
viz. the fixation and subsequent specialisation of larval forms. I have suggested 
the same also to account for the Acarina, which in many respects are simpler 
than I believe the ancestral Arachnid could well have been. 
