200 
A regular epithelium (lining the coelom) was seen to give rise to single 
ova, which, as they became larger, pressed themselves out from the 
regular contour of the epithelium. Had I not known of the existence 
of ganglion cells in the myotomes of later embryos of Raja, in very 
much this position, and had not the various stages in the trans- 
formation of such cells as the two seen in fig. 6 (gl. c’.) into fully ~ 
developed ganglion cells been observed, it might perhaps have appeared 
possible that these two cells were merely primitive ova in a somewhat 
abnormal position!). This supposition can be disposed of; for there 
can be no doubt at all that some of these ganglion cells undergo 
their whole development in situ from cells forming the epithelium 
of the myotome. The importance of this observation, if it possesses 
any, cannot be here discussed. It is well known that KLEINENBERG 
has taken up a very decided position regarding the nature of the 
mesoderm, and that, according to him, it would appear to be largely 
an epiblastic derivative. The work of KÖLLIKER and O. ScHULTZE 
on the epiblastic origin of the mesoderm in Mammals and Amphi- 
bians is also common knowledge. If, in my complete paper, a small 
step can be taken towards disestablishing the right of the mesoderm 
to rank as a distinct germinal layer, I shall feel that once again the 
example of KLEINENBERG has not been lost upon me. 
The explanation of this mesodermal origin of ganglion cells is, of 
course, simple. No doubt many, if not most, of the peripheral ganglion 
cells still find there first origin within the central nervous system, 
wandering thence to their destination. It is merely putting back this 
wandering to a far earlier stage; i. e. before a differentiation into 
segments or myotomes, that gives rise to the opportunity of ascribing 
to them a ,,mesodermal“ origin. 
In an embryo of 43 mm. in length the apparatus may be seen in 
its highest development. In such embryos one may find the elements 
of the larval nervous system massed together in a domain, which 
anteriorly would overlap the region of the thymus; i. e. of the two 
or three posterior gill clefts, while posteriorly or caudalwards it would 
extend as far as the commencement, and for some little distance into, 
the part of the trunk in which lies the genital gland *). 
1) In this connection compare Ricxert, „Über die Entstehung der 
Exkretionsorgane bei Selachiern“ Archiv f. Anat. und Physiol. Anat. 
Abteil. 1888, p. 257. 
2) That part of the apparatus earliest developed is alone here referred 
to. The conditions further caudalwards require renewed investigation. 
