28 
fiEOROE E. N'K'Iinr.ES. 
siib-commissimil organ (ependymal groove) described in 
some detail (’03a)^ Sargent still cdnng (’04) to the idea that, 
at any rate in part, Reissner’s fibre is traceable directly into 
the torus, notwithstanding thnt his original description of 
such relation of the fibre to the torus was based upon an 
erroneons identification of this structure. It is clear from Ids 
descrijitions and also from his text-figure (’00), making 
allowance for the mistaken identification of the parts con- 
cerned, that Sargent actually saw Reissner’s fibre lying in its 
proper position as a single tliread below the posterior 
comndssure (termed by him “torus longitndinalis ”) for 
one half to two thirds of the length of that structure, and 
then breaking up to join the sub-commissural oi-gan (“the 
meiidirane which covers the torus”). In these earlier 
descriptions he nowhere suggests a division of the fibre into 
two main bi-anches. 
In his last paper, however, he shows the fibre composed 
of two main factors — (i) an anterior bi'anch which lies 
beneath tlie posterior commissuie and which is indisputably 
the fibre originally described by him as .solely constituting 
Roissnei’s fibre, and (2) a posterior branch whicli was 
not indicated in any way in either of his prelimi- 
nary papers (’00, ’01), but which is now stated to pass 
above (posterior to) the po.sterior commi.ssnre, and is 
described as emerging into the mesocoel direct!}^ from the 
torus longitndinalis. This posterior branch is said to 
be composed of numerous axon.s, derived, in Teleosts, from the 
cells of the torus, which he therefore homologises with 
the “ Dachkern,” from the cells of which he derives the con- 
stituents of the posterior branch in other types. 
Sai'gent does not, however, call attention to this marked 
discrepancy between his later account and that previously 
given, and, in the absence of any definite statement to the 
contrary, the reader is naturally led to suppose that the 
))osterior branch described in 1904 as directly related to 
the torus is identical with tlie entire fibre erroneously 
described in the preliminary ]>aper (’00) as having the same 
