32 THOMPSON YATES LABORATORIES REPORT 
compounded and pieced togetlier out of several rays or myotoms. Moreover, the boundaries 
between the myotoms do not correspond with the intervals between muscles, nor even with those 
between muscle-groups. Degeneration experiments, which enable one to follow the distribution of 
the individual nerve-fibres of a root, show that in some muscles the number of motor nerve-fibres 
given by a spinal root to a muscle is too small to evoke from the muscle any contraction obvious to 
inspection, for cases occur where a limb-muscle receives three, four, or five motor nerve-fibres 
from a particular nerve root. This I regard as strong testimony to the morphological character of 
the overlap.* 
Another feature of the distribution of the motor fibres of the spinal root to a muscle is the 
remarkable frequency with which it is subject to slight individual variation. In examining 
a series of individuals (cats, monkeys) it is almost rare to meet two consecutive members of the 
series in which the root distribution is not by the degeneration or experimental method 
demonstrably somewhat different. Thus, as instance, I found in some individuals supinator brevh 
innervated from the sixth and fifth cervical nerves; in others, from the sixth and seventh. In the 
former case the innervation of the muscle may be termed '■prefixed'' type, in the latter 
post-fixed.^ \ In- my observations I considered it sufficient to group the individuals into two 
classes, a post-fixed and a prefixed. The absolute segmental level of a muscle is variable over the 
range of nearly a whole segment's length ; the relative segmental position is, however, preserved 
inviolably constant. 
As regards the afferent nerve-fibres of the skeletal muscles, after their existence had been 
proved by the degeneration method, it was possible by the same method to examine their relation 
to the spinal segments. The result of such examination shows that not only are the skeletal muscles 
in many instances, pre-eminently in the limbs, pluri-segmental as regards their motor innervation, 
but that they are so also in regard to afferent innervation. I was able to show J that the afferent 
nerve-fibres distributed to a given muscle arise in the root ganglia of exactly those spinal segments 
whence emerge the motor-fibres for the same muscle. In other words, the sensory nerve-cells 
directly connected with a given skeletal muscle are in any one individual always of the same 
segmental level as are the motor nerve-cells connected with the same given muscle. In so far, 
therefore, the simplest reflex arc connected with a muscle may be expected to lie exactly in those 
segments, whence issue the motor fibres of the muscle ; and is a segmental arc. In the ' knee- 
jerk ' we have evidence of a reflex arc traceable mainly frotn and into vastus niedialis and adjacent 
part of crureus,^ and this affords, as it were, a test case for the above conclusions. It confirms 
them perfectly. It exemplifies them by its narrow local extent, and by the segmentally horizontal 
correlation of the motor and sensory components. 
What was said above to show that the intra-spinal site of the efferent root-cells is at the 
segmental level of the point of emergence of the motor cell-fibres itself, applies to the motor 
nerve-cells of the sympathetic system as well as to those of the skeletal muscles. The sympa- 
* Sherrington, Presidential Address to the Biological Society of Liverfool, 1898. 
f Sherrington, ' Journ. of Physiol.,' vol. xiii, 1892. | ' Phil. Trans. Roy. Soc.,' London, 1896. 
§ Sherrington, ' Journ. of Physiol.,' xiii, 1S92. 
