This page has been proofread, but needs to be validated.
PATHOLOGICAL ANATOMY
15

volutions flat (Wickman). The pia may be hyperemic and very edematous (Daubert). The brain substance may also be hyperemic. The gray matter may occasionally show flecks of red (Redlich). Cerebral changes may be invisible to the naked eye, and microscopic alterations yet be present (Wickman).

The macroscopic changes in the transitional and end stages may be very slight. But if the destruction is extensive, the anterior horn appears shrunken at first and smaller than normal later, after a scar has formed. If only one side be affected, on cross section the cord is asymmetrical; the bundles of nerve fibers adjacent to the affected anterior horn appear smaller than normal and the roots arising from it are atrophied.

Microscopic Changes.—The pia mater. The pia is the seat of a round-celled infiltration. (Plate II, No. 1.) The majority of the round cells appear to be lymphocytes. But another kind of cells (Maximov's polyblast)—which I shall describe more fully in discussing the round cells of the spinal cord—also occurs in relatively great numbers. The Unna Marshalko plasma cells, large lymphocytes and fibroblasts are seldom seen. More frequent than these, though also rare, are typical polynuclear leucocytes with lobed nuclei and neutrophilic granules.

The spread of the infiltration in all of my cases showed considerable uniformity. The infiltration was especially marked in the lumbar and sacral regions of the cord, where it generally implicated the entire circumference. In the upper parts of the cord, it was less severe and could usually be seen only at the entrance of the anterior fissure, chiefly in the anterior fold of the pia. The infiltration was especially pronounced over the front of the cord. But in sections at higher levels, circumscribed areas of marked accumulations of cells also occurred in the pia. Forssner, Sjovall, and Marburg and Strauss observed further that the infiltration was particularly localized to the lower part of the medulla spinalis. In the cases reported by Harbitz and Scheel, the infiltration, although generalized, yet predominated in the lower parts and over the anterior aspect of the cord.

Some special peculiarities of the infiltration of the pia have yet to be mentioned. The contiguous bundles of medullated nerve fibers are rarely implicated. The infiltration of the pia is confined to the area around the vessels. At higher levels, where the