THE LUMBAR ENLARGEMENT OF THE SPINAL CORD. ; 349 
of the loop, thus standing probably in mediate connection with the anterior roots. 
Occasionally these fibres accompany the anterior roots for some distance, in which case 
they appear to share the origin of the anterior roots, either arising from cells or being 
direct continuations of the posterior roots, bending round subsequently, and pursuing 
either an ascending or descending course, finally curving round to re-enter the gray 
substance. A few of the curving fibres found in the anterior columns pass out a 
short distance with some bundle of anterior roots, leaving this soon, and curving 
upwards or downwards to join some other bundle of roots, either above or below, with 
which they pass out from the cord. 
In concluding this paper, I cannot but feel fully aware of its extreme incompleteness ; 
it expresses, however, the constant work of a year, during which time I have had occa- 
sion continually to feel grateful for aid and encouragement from kind friends, in which 
connection I cannot fail to mention the constant interest and kindness of Dr. O. W. 
Holmes, and my great indebtedness for many useful suggestions to Professor Jeffries 
Wyman, without whose aid I should hardly have begun this work, and certainly 
should not have carried it through. : 
The accompanying figures have all been drawn by myself, by means of the camera 
lucida; and although much interesting detail is necessarily omitted, I have endeavored 
to make them faithful representations of the most important facts, my constant effort 
having been to give too little detail rather than too much. I have etched them on 
copper myself, believing that microscopic drawings are too apt to lose much of their 
truth even in the hands of the most careful artist, and that a faithful, though some- 
what rough transcript of the original drawing, made by the observer himself, must be 
of greater value than a more highly finished copy. e 
Conclusions. 
1. That true nerve-cells exist only in the gray substance, these cells being connected 
by their processes in more or less definite groups; but not probably, as some authors 
maintain, so as to form an uninterrupted chain from brain to filum terminale. 
2. 'That the anterior roots arise partly from nerve-cells, another portion being directly 
continuous with the posterior roots. A part of the posterior roots also enter cells. 
3. The roots may therefore be divided into three classes ; — such a division does not, 
however, imply any functional difference. 
(1.) Anterior and posterior roots which arise from or terminate in anterior or poste- 
rior cells. 
VOL. VIII. 45 
