12 
Proceedings of the Royal Society 
believe that it could first arise from tbe intestine, and afterwards 
pass into its present position. The notochord must have arisen 
phylogenetically from the mesoblast between the neurochord and 
the aorta. Some of my sections of herring embryos show the noto- 
chord distinctly marked off from the hypoblast, and continuous 
with the neurochord. This is not to be interpreted as showing that 
the notochord was derived from the neurochord, a skeletal structure 
from an epiblastic. The reason why the notochord is at first in 
close relation with the hypoblast, or with the neurochord, is that its 
development in the embryo has come to take place at so early a 
stage that the fusion of the three layers along the line of the primi- 
tive blastopore has not disappeared before the notochord appears. 
Balfour says, that in Elasmobranchs it is difficult to ascertain 
whether the notochord is derived from the hypoblast or is a central 
column of mesoblast cells. We thus get back to the old view, that 
the notochord is homologous with the three giant fibres beneath the 
nerve-chord in the earth-worm; both are mesoblastic structures, 
developed in the same position for the purpose of supporting the 
nerve-chord. It has been suggested that the typhlosole in the 
earth-worm represents the notochord ; as I have shown this is 
impossible ; the typhlosole reappears in the spiral valve of the in- 
testine of Elasmobranchs and other fishes. The fact that the noto- 
chord stops short just behind the infundibulum is fully accounted 
for by my theory that the latter structure is the primitive mouth. 
The Vertebrate Eye . — The peculiarity in the development of the 
vertebrate eye, as compared with that of Invertebrates, has long 
ceased to be so great a mystery as it was to earlier morphologists. 
In his address to the Department of Anatomy and Physiology of 
Section D, at the British Association meeting at Swansea in 1880, 
Balfour pointed out that the retina was formed from the floor of 
the brain, that is to say, from the same portion of the primitive 
epidermis which formed the central nervous system. In a similar 
way the rhabdoms or retinal elements in the eyes of Arthropods are 
formed from the supraoesophageal ganglia. In his little book on 
Degeneration in the Nature series, Prof. Lankester has inferred 
from the consideration of the cerebral eyes of Ascidians, that the 
original vertebrate ancestor was transparent, and had eyes on the floor 
of its brain ; that when the animal became opaque the eyes gradually 
