No. 388.] CORRESPONDENCE. 361 
tion of this problem any easier, and in the end leads to hope- 
less confusion. It will, therefore, be interesting and profitable to 
consider some of the difficulties into which Mr. Gaskell is led in 
his attempts to solve this problem by the novel method of com- 
paring the dorsal surface of an arthropod with the dorsal surface 
of a vertebrate. 
It may be stated incidentally that Mr. Gaskell adopts, without 
acknowledgment, the same lines of argument in reference to many 
homologies between vertebrates and arthropods that were used in 
my first paper on this subject in the Quarterly Journal. This is 
notably the case in regard to the paleontological evidence, the 
relation of the endosternite of arachnids to the cartilaginous 
cranium of vertebrates, and to the causes for the disappearance 
of the old mouth in the concentration of the thoracic neuromeres 
around the arachnid cesophagus, although these facts are quite 
inapplicable to his theory. 
Briefly stated, Gaskell maintains (Journ. of Physiol., 1889, p. 191) 
that the nervous system of vertebrates is composed of two essentially 
different parts: first, a preéxisting, non-nervous tube, consisting of 
the epithelium of the canalis centralis and cerebral vesicles, and the 
various supporting elements derived from it; and, second, the true 
nervous elements, consisting of a ‘bilateral chain of ganglia con- 
nected together by means of longitudinal and transverse commis- 
sures.” The infolding of the medullary plate of vertebrates shows 
us, he maintains, the simultaneous development of two different 
organs, the one the nervous system, and the other the tube of 
supporting tissue, p. 193. This tube of supporting tissue, “ which 
is not nervous and never was nervous,” and which is coextensive 
with the canalis centralis and the cerebral ventricles, Gaskell regards 
as the remnants of the alimentary canal of a crustacean-like ancestor. 
The ventral cord and the supracesophageal ganglia of the crusta- 
cean ancestor have in vertebrates fused with and grown around the 
old alimentary canal to form the true nervous elements of the 
spinal cord and brain. No reversal of surfaces is called for by this 
transformation, for the ventral surface of an arthropod is regarded 
as homologous with the ventral side of a vertebrate.... He 
maintains that the one reason why they (the champions of the 
origin of vertebrates from the appendiculata) have not been able 
to make any real advance in their views, has been the difficulty of 
accounting for the altered relations of alimentary canal and ner- 
vous system in the two groups. His theory “ solves this difficulty,” 
