SENSE ORGANS OF AMIURUS. 263 
The R. lateralis vagi which supplies the sense organs of the canal is 
not situated in the subcutaneous tissue beneath the canal, but a little 
distance inwards between the two masses of muscle, a branch being 
detached to pass outwards to each nerve-hillock. In transverse 
sections through the canal, it is obvious that it is situated between 
the epidermis and the stratified fibrous layer of the corium, being 
lodged in what is elsewhere the pigmentary layer of the corium, 
although the pigment is practically absent in the neighbourhood of 
the canal. The epithelium of the canal which is quite low, except 
where it is transformed into the neuro-epithelium of the nerve. 
hillock, is continuous at the pores with the surface epithelium of the 
skin. An exceedingly delicate connective tissue surrounds the 
epithelium, separating it from the proper wall of the canal, which is 
formed in the neighbourhood of the pores of a dense connective 
tissue whose elements are disposed radially to the wall of the canal, 
but in the neighbourhood of the nerve-hillocks, and indeed for the 
greater part of the canal between the pores, by a much thinner layer 
of osseous substance, so disposed as to form a complete tube for the 
greater part of its course, but less complete towards its ends. No 
bone corpuscles are present in the osseous wall of the canal, as is 
also noted by Leydig and Bodenstein for the forms described by 
them. I am unable to identify the above-mentioned dense connec- 
tive tissue with cartilage as Bodenstein does, the corpuscles are quite 
similar to connective tissue corpuscles, and there is no matrix stain- 
ing in Bismarck brown, as is the case even in cartilage which has a 
minimum of intercellular substance. Separating the dense wall from 
the surrounding tissues is again a layer of reticular tissue belonging 
to that which I have above spoken of as the pigmentary layer of the 
corium. 
The lateral canal of the adult is approximately 2mm. in transverse 
diameter ; in young specimens of two inches in length, hardly one- 
third of that. 
To study the course of the mucous canals in the head a series 
through young forms is most convenient, although approximately the 
direction of the canals may be seen also from the pores. (Figs. 4, 
5, 6.) The pores do not open directly into the canals of the head 
as they do into that of the lateral line, but by longer or shorter 
tubes—-a circumstance noted also by Bodenstein for Cottws—and con- 
