A Case of Glioma of the Lower Cervical 
Region of the Spinal Cord Producing a 
Total Transverse Lesion, in which there 
was Spasticity of the Lower Limbs and 
Persistence of the Deep Reflexes. 
/. A Study of the Tumor and Some of the Phenomena 
Produced Thereby. 
By W. H. HUDSON, M.D., 
OF LAFAYETTE, ALA. 
II. A Study of a Peculiar Form of Dege7ieration (Degen- 
era tio Micans) Met with in this Case. 
By LEWELLYS F. BARKER, M.D., 
ASSOCIATE PROFESSOR OF ANATOMY, JOHNS HOPKINS UNIVERSITY. 
III. Some Points in the Pathology of Syringomyelia. 
By SIMON FLEXNER, M.D., 
PROFESSOR OF PATHOLOGICAL ANATOMY, JOHNS HOPKINS UNIVERSITY. 
FROM 
THE AMERICAN JOURNAL OF MEDICAL SCIENCES, 
June, 1899. A CASE OF GLIOMA OF THE LOWER CERVICAL REGION OF 
THE SPINAL CORD PRODUCING A TOTAL TRANSVERSE 
LESION, IN WHICH THERE WAS SPASTICITY OF 
THE LOWER LIMBS AND PERSISTENCE OF 
THE DEEP REFLEXES. 
I. A STUDY OF THE TUMOR AND SOME OF THE PHENOMENA 
PRODUCED THEREBY. 
By W. H. Hudson, M.D., 
OF LAFAYETTE, ALA. 
II. A STUDY OF A PECULIAR FORM OF DEGENERATION (DE- 
GENERATIO MICANS) MET WITH IN THIS CASE. 
By Lewellys F. Barker, M.D., 
ASSOCIATE PROFESSOR OF ANATOMY, JOHNS HOPKINS UNIVERSITY. 
III. SOME POINTS IN THE PATHOLOGY OF SYRINGOMYELIA. 
By Simon Flexner, M.D., 
PROFESSOR OF PATHOLOGICAL ANATOMY, JOHNS HOPKINS UNIVERSITY. 
I. 
In August, 1890, Mr. \V. C. T., white, aged fifty years, an intelligent 
and well-to-do farmer, consulted me on account of difficulty in walking 
and pain, which he located in his spine, from the upper dorsal region 
downward. He also complained of burning in both legs and feet. 
His family history was good, no member of his family having died 
with tuberculosis or cancer. About six months before seeing me he 
had fallen off his horse backward, striking on his shoulders and neck. 
He attributed his sickness to this accident. 
He was a well-nourished man, with no deformity of the spine and 
no well-marked point of local pain. He had no history of syphilis. 
His knee-jerks were considerably exaggerated, and he swayed markedly 
when standing with his eyes closed. 
I gave him iodide of potassium in increasing doses and bichloride 
of mercury, which he took for about ten weeks, his condition, in the 
meantime, growing slowly and gradually worse. 
In November, 1890, he passed out of my care, and the history of his 
illness from that time to December, 1896,1 have obtained from the physi- 
cians who attended him and from the members of his family. During 
December, 1890, he had a severe attack of what his physicians called 
dysentery, and afterward his abdomen was greatly swollen for a long 
time. After this attack he was able to walk only with the aid of crutches. 
His bladder gave him trouble, and he also had more or less difficulty in 
evacuating his bowels. During the early part of 1891 a girdle sensation 2 
developed, which was located somewhere in the region of the umbilicus. 
From the summer of 1891 until his death he was never able to walk 
even with the aid of crutches. During 1892-1894 he had several 
attacks of dysentery, and at these times passed pieces of the mucous 
membrane of the bowels. His abdomen much of the time was swollen. 
There was progressive paralysis of motion, the right leg becoming par- 
alyzed some time before the left, and the patient thought that he had 
sensation longer in his left than in his right leg. I could get no reli- 
able information as to where the paralysis of motion or sensation began. 
At times he used a catheter to draw his urine, but there appears to be 
no history of well-marked cystitis. His bowels usually moved every 
day or two, occasionally requiring the aid of an enema. 
From the early part of 1895 to the summer of 1896 his general 
health, according to his statement, was excellent; his appetite was 
good; he was cheerful; did not suffer much pain and gained in flesh 
until he weighed nearly two hundred pounds. He had never at any 
time suffered the excruciating pain common to spinal tumors. At times, 
however, there was pain in the spine and a marked circle of pain around 
his body. This girdle of pain was always equally marked on both sides 
of the body, and slowly advanced upward until it reached a point above 
the nipples. All sensation and motion did not disappear until the latter 
part of 1895. The patient thought that after this time he could make 
no voluntary motion and feel nothing below the nipples. 
He was troubled a great deal with what he called “ jumping of the 
legs,” the merest touch of the bedclothes frequently putting his legs to 
jumping, this “jumping” keeping up for some time. When bathing 
his feet the same trouble would be experienced, and he frequently 
kicked the tub over or splashed the water about. 
During the period of his so-called good health he passed his urine 
at intervals of four or five hours, and his bowels moved every day or 
every other day, only occasionally requiring the aid of an enema. 
His legs were more or less rigid, and at times there was a tendency 
to flexion ; but for some time past they had remained straight and 
spastic, with the feet somewhat extended. 
There had at no time been bed-sores. 
There was one phenomenon about which he was very emphatic—the 
effect that strychnine, which was frequently given him, produced. Fre- 
quently during the course of his disease he suffered severely with burn- 
ing in his body, legs, and feet. Strychnine invariably caused such 
increase of this burning and such increase in the “jumping” of his 
legs, though it caused no other inconvenience, that he refused to take 
it further. 
In November, 1896, he made a visit to his daughter, who lived sev- 
eral miles away. While on this visit he was taken with vomiting, 
which he attributed to the eating of cheese. He returned home, and 
I saw him during the early days of December. 
Notes taken at my first visit: 
Patient is pale and greatly depressed in spirits. He has no fever, 
but his pulse is about 110 beats per minute. He is still vomiting occa- 
sionally, and says he can find no food which agrees with him. His 
abdomen is slightly swollen, but much less than it has been for a week 
past. His breathing is entirely diaphragmatic, and he suffers from 3 
difficulty of breathing and a troublesome cough, which occasionally 
brings up a small quantity of frothy mucus. There is no rigidity of 
the back, and he usually suffers no great pain, but pain is produced 
when he turns his neck. In a darkened room his pupils appear smaller 
than normal, and they react indifferently to light. The palpebral 
fissures are somewhat narrowed. 
Sensation. There is total absence of tactile, algesic, and thermic sen- 
sibility below the third rib in front and a little below the spine of the 
scapula behind, on the left side; on the right side the insensibility 
reaches a higher level. There is no dissociation of sensation at any 
point. Subjectively, there is some pain down his right arm and over 
the lower posterior portion of the neck when he turns his head. There 
is hyperalgesia over the top of the right shoulder and about both sides 
of the neck. All sensations are obtunded over the uluar side of the 
right arm, forearm, and hand. 
Trophic Changes. The legs are somewhat rigid, and their muscles 
appear firm and not wasted. The hip-joints, knee-joints, and ankle- 
joints are stiffened, but can be flexed, to some extent, by using force. 
The skin is not especially dry, scurfy, or glossy, and appears quite 
normal. There are no bed-sores, nor have there been any during the 
whole course of his sickness. The interossei between the little and ring 
fingers and between the ring and middle fingers of the right hand and 
the hypothenar eminence are somewhat atrophied. 
Condition of the Reflexes. When the skin over any portion of his 
legs is pricked there is a motion of flexion. When the soles of his feet 
are tickled there is a motion in the toes and a tremor extends up the 
legs. When the toes are sharply lifted toward the knees a violent ankle- 
clonus begins, which lasts while the tension is kept up. The knee-jerks 
are well marked in both legs and considerably exaggerated in the left. 
I failed to elicit either the cremasteric or epigastric reflex. A digital 
examination of the rectum reveals the presence of well-marked contrac- 
tion in the sphincter ani. (I did not insert a catheter into his bladder, 
and I was never able to get any of his urine for examination.) There 
is no deformity or rigidity of the spine, but over the upper dorsal 
region the percussion note is increased in resonance, and over the lower 
cervical region there is increase in dulness. 
Diagnosis. A tumor of the spinal cord, probably intradural and, I 
fear, intramedullary, but on account of the persistence of the reflexes 
and the slow growth of the tumor I entertain the hope of operative 
relief should the patient gain sufficient strength. 
The patient lived some distance from me, and during the next six 
weeks I only saw him four times, the last visit being made seven days 
before his death. At this time the reflexes were still present, but 
weaker. There was a bed-sore beginning behind the left great trochan- 
ter, and on the dorsal surface of the left foot there was a spot of roughly 
elevated, bluish skin, nearly as large as the palm of the hand, which 
had the appearance of a dried blister. During the last few days of his 
life his right arm was paralyzed to motion, and he suffered intense pain 
in both arms. 
He died January 17,1897. The exact cause of death I was not able 
to ascertain. 
Four hours after death I was allowed to open only the spinal canal, 
as would have been done had an operation been performed. 4 
The spinal cord, from the fourth cervical vertebra to the third dorsal 
vertebra, was removed. This portion of the cord is shown in Plate I. 
The upper five centimetres of the specimen—the part containing the 
enlarged tumor-formation—measured at the broadest point, laterally, 
21 mm.; at the same point antero-posteriorly, 11 mm. The enlargement 
tapered gradually downward from the centre until, at the lower limit of 
the tumor, the measurements were about the same as those of a normal 
spinal cord. Immediately below the tumor, for about three centimetres, 
the cord was a softened mass; for a short distance it was not much 
larger than a small goose-quill, and bent of its own weight in any 
direction. The cord was placed at once into Muller’s fluid, and on the 
next day was cut into sections about one centimetre long. The tumor 
was then plainly seen. It was a little darker in color than the cord, 
but both cord and tumor were stained with Muller’s fluid, and the 
exact color of the tumor could not be made out. At some points the 
growth was about the consistency of normal spinal cord; at others, 
especially about the centre, the tissues were somewhat of a spongy 
appearauce, while at other points cavities were plainly seen. The upper 
end of the specimen revealed two cone-shaped points, each situated in 
the region of the posterior horns of gray matter; these appeared to be 
the upper limit of the tumor. The first cut through the upper part of 
the specimen passed through an oblong, smooth-walled cavity measuring, 
laterally, 8 mm.; antero-posteriorly, 4 mm.; longitudinally, 5 mm. It 
contained the remains of a firm blood-clot the size of a small pea. 
The third cut through the tumor, passing about its middle, revealed 
a lateral slit 10 mm. long. At this point the tumor appeared to occupy 
the entire cord, with only membranes surrounding it, which fell away 
from the sides of the tumor as the knife passed through. 
The lower portion of the tumor presented an appearance of two cones 
lying side by side, with their apices jutting into the softened cord below. 
The softened cord appeared entirely disorganized, except the central 
stem of gliomatosis which passed through it. The lower limit of the 
specimen retained the normal, cord-like shape, with the central stem of 
gliomatosis occupying the position as indicated in Plate I. 
Microscopical Examination. Section 1, Plate I. The lateral meas- 
urement of the cord at this point is 20 mm.; the antero-posterior, 11 mm. 
The anterior gray commissure has disappeared, and the tumor is plainly 
seen occupying a large portion of the cord. The mass of tumor extends 
right up to the direct pyramidal tracts, separated from them by a clear 
space and a few degenerated nerve-fibres only; these form a broken 
boundary running along the anterior border of the tumor. Nerve- 
fibres running in the same direction are seen here and there around 
the entire periphery of the tumor. To the left of the anterior commis- 
sure, just within the tumor mass, is a microscopical cavity of consider- 
able length and of regular shape. This has a lining of cylindrical 
epithelium; at each end of the cavity the epithelium is irregular, and 
appears to have straggled from the line. The cavity is perhaps the 
remains of the normal central canal of the spinal cord. Anteriorly 
the rim of normal cord measures 2 mm., but it gradually becomes 
thinner posteriorly until all nerve-fibres disappear. 
In the anterior rim of the cord many of the nerve-fibres are normal. 
The irregular outlines of the anterior horns of gray matter are also to 
be seen, with many ganglion cells, some of them more or less degen- 5 
erated. Posteriorly the nerve-elements are more degenerated. The blood- 
vessels in the remnant of the cord and in the pia mater do not appear 
much changed. Conspicuous in the tumor is the great number of blood- 
vessels. Most of them, especially in the periphery, have immensely 
thickened walls, and in some instances the lumen is a mere slit or is 
even entirely obliterated. In some the vasa vasorum are yet to be 
seen. As the centre of the tumor—the region surrounding the main 
cavity—is reached there are found a few large, thin-walled bloodvessels 
(veins ?) filled with blood-corpuscles. There are also in this locality 
numerous smaller vessels, with thick, homogeneous walls, many of which 
appear as solid rods. 
Besides the bloodvessels the tumor is made up of cells and fibres. 
The cells are round, oval, and pyriform in shape, and in many places 
are arranged in a radial manner around the bloodvessels. They are 
most numerous in the periphery of the tumor, and toward the centre 
diminish in number. They are not, however, always regularly dis- 
tributed, but are massed more abundantly in places and lessened in 
others. The fibres are also everywhere to be seen—fewest where the 
cells are most numerous, most where the cells are fewest. (For fur- 
ther description of the fibres and cells see Dr. Flexner’s account.) 
In Section 1 two cavities are to be seen—the larger centrally located, 
the smaller roughly S-shaped and located to the right. As the brink 
of the larger cavity is reached the cells become very much diminished 
in number, and the fibres are very abundant. Here and there this 
cavity is lined with cylindrical epithelium, and when this is the case 
these epithelial cells are nearly always reinforced by lines of cells 
immediately in their rear. Where there is no epithelial lining to the 
cavity the border is made up of neuroglia fibres, many of them pro- 
jecting into the cavity in a very ragged manner. At various points 
around the edge of the cavity the glia fibres appear thickened and 
nodular. There are also many peculiar bodies; some are spindle- 
shaped, others horn-shaped, and some varicose, while others present 
small, nodular formations. Within this cavity there was a blood-clot 
as large as a small pea. 
The surroundings of the smaller cavity differ from those of the larger, 
inasmuch as the cells around the smaller are fewer and the fibres and 
cells are much changed. There are also fragments of epithelial lining 
in this cavity, and here the epithelial lining cells are reinforced by cells 
immediately in their rear. The neuroglia fibres making up the border 
of this cavity show more marked changes than those forming the border 
of the larger cavity. Just behind this cavity the peculiar particles 
mentioned above are very numerous and strike the attention when the 
specimen is examined. 
Section 2, Plate I This section is taken about 1 centimetre lower 
than Section 1, and measures 21 mm. laterally and 11 mm. antero- 
posteriorly. There is a continuation of the strip of cord which appears 
in Section 1. The anterior columns are to be seen, the anterior median 
fissure being displaced to the left. On the left a few ganglion cells are 
lying along the border of the tumor. The tumor has displaced the 
cord forward and outward until, posteriorly, all the nerve-elements have 
disappeared, and laterally they are very much degenerated, the strip of 
cord becoming thinner posteriorly until it disappears. 
The tumor is very plainly circumscribed, and anteriorly and on the 6 
side has a boundary of degenerated nerve-fibres running along its 
border. The stnicture of the tumor as to bloodvessels, cells, and fibres 
is entirely similar to that of Section 1. In the centre of the tumor is 
a cavity of irregular shape, filled here and there with fibres and a few 
cells, these being apparently the debris of the tumor as it appears in 
the immediate surroundings. There also appears here and there in the 
wall of the cavity and in the surrounding tissue broken lines of epithe- 
lial cells and some small, broken cavities with epithelial lining. In 
the immediate vicinity of this cavity bloodvessels (veins ?) with thin 
walls, all filled with blood-corpuscles, are numerous. Many of these 
vessels are not round, but are irregular shaped—some as if two or three 
had been united and the partitions removed. The smaller arteries, 
with their slit-like or closed lumina, are also numerous. This cavity is 
evidently in a state of active formation. In the posterior border of the 
tumor there are several cavities, very irregular in shape, with a lining of 
exquisite cylindrical epithelium. 
Section 3, Plate I (from a section one centimetre lower than Sec- 
tion 2.) Here the tumor reaches its greatest lateral measurement—21 
mm., measuring antero-posteriorly 10 mm. The remains of the cord 
consist of a thin rim, less than 1 mm. thick in front and disappearing 
as the posterior border is approached. A few scattered ganglion cells 
are still to be found along the left anterior border of the tumor. The 
nerve-elements everywhere show marked degeneration. The cavity 
and its surroundings are very similar to those in Section 2. Here and 
there are fragments of cylindrical epithelium. The rod-like hyaline 
arteries are very numerous, and the thin-walled vessels (veins?) are 
filled with blood-corpuscles. In places these thin-walled vessels have 
ruptured and the blood has been extravasated among the cells and fibres. 
In the posterior region of the tumor there is a small, irregular cavity, 
lined here and there with a broken layer of epithelium. Near the 
centre of the posterior border there is a larger cavity, extremely irregu- 
lar in shape, with thick-walled bloodvessels, standing upright, surrounded 
by radially arranged fibres and cells and covered by a beautiful layer 
of cylindrical epithelium. These formations stand out into the open 
space like Grecian columns, and present a very striking picture. 
Section 4, Plate I (taken from about 1 centimetre below Section 3). 
At this point the tumor has a more rounded shape, measuring laterally 
18 mm.; antero-posteriorly, 11 mm. The medullated nerves in the 
remnant of the cord are reduced on the left to a very small space on 
the anterior border, and on the right to a small space near the anterior 
fissure. There are no ganglion cells to be seen. Further out on either 
side, especially on the right, the degenerative changes are very marked. 
The structure of the tumor is the same as at the other points described. 
The conditions in the central cavity are also the same as in the other 
sections. On the left side there is a long, irregular cavity, which at 
some points has a lining of cylindrical epithelium, and is connected 
with the cavity by the disintegration of tissue. Posteriorly we have 
the same epithelium-lined cavity as described in Section 3, but the 
beautiful Grecian columns have lost their contour. Far to the right 
there is a roughly star-shaped cavity lined with cylindrical epithelium. 
Section 5, Plate I. (taken about 1 centimetre lower than Section 4). 
At this point the tumor and remnant of the cord measure laterally 15 
mm.; antero-posteriorly, 9 mm. The thin rim of cord, thicker in front 7 
than behind, more or less degenerated, extends nearly around the tumor, 
failing by a short space to connect dorsally. There are along the ante- 
rior border of the tumor a few degenerated ganglion cells to be seen. 
These do not appear to retain their processes. The nervous elements 
throughout the strip of cord appear much degenerated. The tumor at 
this point has undergone less degenerative changes than in any section 
heretofore described. There are no cavities which have had their origin 
in retrograde changes. Here also there are thin-walled vessels (veins ?). 
The arteries with thickened walls are very numerous, and in nearly 
every case they show the radial arrangement of the cells and fibres; 
indeed, it is in this region of the tumor that this is most characteristi- 
cally shown. It would appear that this radial arrangement of the cells 
and fibres is most commonly seen in that part of the tumor which yet 
preserves its natural structure, and must have existed around most or 
all of the arteries at some time in their growth. In that portion of the 
tumor which was between Sections 4 and 5 the preformed or epithelial- 
lined cavities were most numerous; at some places the structure of the 
tumor here might have been called cystic. 
The tumor at the point from which Section 5 was taken showed very 
few signs of degeneration, although at several points around the epithe- 
lial-lined cavities the fibres are greatly increased over the cells, and 
these areas are paler than the other portions of the tumor. 
Section 6, Plate I. This section was taken 1 centimetre below Sec- 
tion 5, and just a little way above the lower limit of the true tumor 
mass. Here the location of the apices of the tumor is seen, also the 
normally located central canal of the cord. The cavities contain epithe- 
lial lining and the remains of hemorrhages. 
Below the true tumor for about 3 centimetres the cord, with the 
exception of the main stem of gliomatosis, was softened, the cord 
elements remaining as a mass of debris only. For a short distance 
below the true tumor, and apparently continuous with its two points, 
are two stems of gliomatosis, one soon disappearing, the other acquiring 
a central position and is continuous throughout the remainder of the 
removed portion of the cord. (Plate I.) In the main stem of the 
central gliomatosis, a short distance below the true tumor, there is 
situated well anteriorly and displaced to one side a deposit of cells, 
which is evidently the remains of the normally located central canal of 
the cord. The central area of gliomatosis in the softened part of the 
cord hardened well in Muller’s fluid, and the fragments of the cord 
were held around it by the membranes. 
Near the centre of the stem of gliomatosis, but distinctly separated 
from the original central canal, running from near one side to near the 
other side of the longest diameter of the stem, is a narrow slit, in some 
places not perfectly formed, in others an actual cavity, to be seen plainly 
with the unaided eye. In the stem of gliomatosis the bloodvessels are 
in striking contrast to the bloodvessels in the true tumor-growth. Their 
appearances in the true tumor have been described; in the area of 
gliomatosis they are not remarkably numerous, nor do their walls show 
any special changes. In the periphery of the area of gliomatosis the 
appearance is quite similar in all parts of the section ; but as the centre 
is approached numerous bodies, which are to be described later, are to 
be seen. These bodies are located as seen in Fig. 2 and Fig. 3, Plate 
III. They present many different forms. Some of them are much 8 
smaller than the unchanged cells making up the true tumor; some 
resemble thickened glia fibres (fibres in the sense of Weigert); some 
of them are larger than ganglion cells and are like them in many 
respects. They are round, oval, pyriform, club-shaped, nail-shaped, 
elongated, bulbous at one end, and gradually taper to a point not larger 
than the glia fibrils, being largest at the middle and tapering to a point 
at each end. They are screw-shaped, horn-shaped, twisted ; many of 
them are varicose, and many, when stained, resemble granular tube- 
casts. These bodies are also to be found in the tumor proper, especially 
around the posterior border of the smaller cavity. (Fig. 1, Plate II.) 
The most characteristic of these formations are not numerous at the 
very brink of the cavities (although some are seen lying free in the 
cavities, and many can be found making up the cavity walls), but just 
a little way removed. In some of the specimens examined they are 
not seen at the points where the cavities yet retain their epithelial lining. 
From the resemblance that a few of these bodies have to ganglion 
cells I suspect that they have been mistaken for ganglion cells by some 
observers. Turner and Mackintosh (Brain, 1896, Parts II. and III.) 
describe a new growth with glia formation in the spinal cord. They 
give a photo-micrograph of a section containing a growth of glia, and 
in this they describe certain formations as most probably ganglion cells. 
The condition pictured by them is very similar to the condition existing 
in the stem of gliomatosis in the case here described, and I should think 
that the bodies which they have described as ganglion cells are the same 
as those present in my case. Turner and Mackintosh consider their case 
to be an example of the condition which has been described as neurog- 
lioma ganglionare. Dr. W. Rosenthal (Zeigler’s Beitrdge, 1898, Band 
xxiii, Heft 1) describes a neuro-epithelioma of the spinal cord with 
central gliomatosis. He found and describes bodies similar in shape and 
location to those discovered in the case here presented. From an exam- 
ination of some of Dr. Rosenthal’s specimens, and from his description 
of the formations, I am of the opinion that the bodies found in his case 
and in mine are entirely similar in nature, although differing in size 
and shape. In a letter to me Dr. Rosenthal expresses the same opinion. 
I wish to thank him for sending me specimens from his case and for 
writing me his views of the peculiar formations. He is of the opinion 
that they represent a peculiar degeneration of the glia fibres (fibres in 
the sense of Weigert), and suggests the name “ Ivolbige Degeneration.” 
From a study of the bodies seen in my case I reached the opinion that 
some of them were changed and degenerating bloodvessels, and were, 
perhaps, some changed glia fibres (glia fibres in the sense of Weigert); 
but that a vast majority of them represented a degeneration existing in 
the bipolar cells, of ependymal type, which are found throughout the 
entire mass of the growth. I here merely wish to note also that though 
these bodies are not conspicuous as lining the cavities or making up the 
tissue which lines the cavities (though they may be found lying free in 
the cavities and many in the lining of the cavities), that they, or, 
rather the degeneration which is going on in them, do perhaps take 
some part—probably an important part—in the production of cavities, 
and especially of cavities which form in the new growth of glia. This 
statement receives some confirmation from the fact that these bodies 
are most numerous in the border of the smaller cavity (Plate II, 
Fig. 2). The tissue surrounding the smaller cavity in the figure just 9 
mentioned is more of the nature of the tissue found in the stem of 
gliomatosis below the mass of the tumor than any other tissue to 
be found in the tumor proper. There are other cavities in the tumor 
proper which appear to have been produced by the hemorrhages from 
the thin-walled bloodvessels (veins?), the result being a disintegra- 
tion of the cells and fibres which make up the tumor. The general 
lack of nutrition produced by the obliteration of the arteries also is, 
perhaps, an important factor in the production of these cavities. The 
existence of obliterating arteries and large, thin-walled vessels forming 
lakes of blood evidently produced a state of the blood-current which 
favored the disintegration of the central region of the tumor. Many 
cavities formed by the disintegration of tissue are connected with pre- 
formed epithelium-lined cavities, and it would be somewhat difficult 
to say in which process the cavity formation had its origin—that is, 
whether the cavity formation began first in the degeneration of the 
cells and fibres around the preformed cavities or in the disintegration 
of the tissues from nutritional changes. 
After studying for awhile the peculiar degeneration in the tumor 
elements, which I have described in a very superficial way, I came to 
the conclusion that the subject was of sufficient importance to demand 
the hand of an expert, and after consultation with Dr. William H. 
Welch, Dr. Lewellys F. Barker kindly did the work, an account of 
which is here appended, and for which I am very grateful. I wish 
especially to thank Dr. Simon Flexner for many kindnesses and his 
contribution, which has been in my hands for several months. Also 
I wish to say that my study of the degeneration in the glia-elements 
was made during the summer of 1897, several months before the publi- 
cation of Rosenthal’s paper. I desire to thank Mr. Max Brodel, the 
artist, for his interest in the work and his excellent drawings, which 
have been well reproduced by the heliotype process. 
Origin of the Tumor. It will be observed that in the history 
here reported the patient attributed the disease to trauma—i. e., falling 
backward off his horse. He had been in perfect health up to the time 
of this accident, and only a short time—a few weeks—afterward he 
noticed the first symptoms of spinal-cord disease. He was conscious, 
however, from the time of the fall that his back was injured in the 
lower cervical region. 
In other cases of tumor formation in the spinal cord there has been 
history of trauma; and it is quite probable that injuries may start into 
activity the latent embryonic elements from which these pathological 
formations have their origin. 
For further details of the histology and histogenesis of the condition 
I refer to Dr. Flexner’s account. 
As the patient whose case is here reported came into my hands for 
surgical aid, a few words on the surgical aspect will not be amiss. 
Starting with the assumption that spinal-cord tumors (I here exclude 10 
tubercular and syphilitic neoplasms) offer no hope of relief through 
any other therapeutic means, they appeal alone to the aid of surgery. 
Tumors of the spinal cord are rapidly and surely fatal, and when, 
according to the cases collected by Starr, such a large per cent, can be 
reached and removed by the surgeon, it becomes our imperative duty 
to furnish these unfortunate sufferers with all the relief that surgery 
affords; but here I do not refer, of course, to intramedullary tumors, 
for one could hardly conceive of the successful removal of a tumor so 
placed, but a circumscribed intramedullary tumor may yet become of 
interest to the surgeon in other ways than from a purely diagnostic 
point of view. It is evident that in exceptional cases, either from lack 
of accurate history of the case or from symptoms so nearly alike in 
each condition, there may be great difficulty in deciding, even with a 
reasonable degree of certainty, whether the tumor is intramedullary 
or extramedullary. Fortunately for the surgery of the spinal cord, 
which was so brilliantly inaugurated by the case of Gowers and Hors- 
ley, these mystifying cases of intramedullary tumors are not very fre- 
quent. 
Operations have been performed not intentionally (as would likely have 
been done in the case here reported had the patient’s strength permit- 
ted) in cases of intramedullary growth. Some of these operations are 
not to be regretted either by the unfortunate sufferers or the surgeons ; 
for as long as it is justifiable to prolong life and lessen suffering, even 
in hopeless cases, any reasonable therapeutic agent is advisable. 
It should be borne in mind that in many of the cases of intramedul- 
lary tumors belonging to the class of the one here reported the de- 
struction of the nerve-fibres is a slow process, and would probably be 
rendered much slower but for the resistance made by the encompassing 
bony walls of the spinal canal. 
Operation for relief of intracranial pressure in tumors of the brain 
is not only justifiable, but in some cases is even attended with a won- 
derful improvement of the general condition of the patient; and not 
only do the symptoms produced by the pressure disappear, but there 
may be a recession of the growth itself. 
Naturally the anatomical and physiological, and even the mechan- 
ical, conditions existing in the spinal cord and its envelopes—membranous 
and bony—are different in many respects from these conditions as they 
exist in the brain. 
Bearing in mind the absolute necessity for the integrity of the con- 
ducting fibres in the spinal cord and the role played by the bony canal 
in the destruction of these fibres in intramedullary growths (gliomata), 
it is not difficult to conceive of cases when life may be prolonged and 
made more comfortable by an operation done solely to relieve pressure. 
Then, when the diagnosis is impossible between an intramedullary 11 
and an extramedullary tumor, the rule should be to operate; and when 
there is reasonable certainty that the growth is intramedullary and 
circumscribed, and that the relief of pressure would prolong and make 
life more endurable, I see no reason why an operation for the relief of 
the pressure (thus preserving for awhile longer fibres in the spinal cord) 
would not be quite as justifiable as an operation made to relieve intra- 
cranial pressure and its attending devastations. 
Tiie Reflexes. From a careful examination of the tumor and the 
softened cord below there can be no doubt as to the lesions having pro- 
duced a total transverse division of the cord ; and this destruction of 
continuity must have occurred at least some months before the death 
of the patient. 
Until within the last few years it was considered a well-established 
fact in the pathology of the nervous system that total transverse lesions 
of the spinal cord above the lumbar region, where the lumbar enlarge- 
ment was not implicated, produced an increase of the deep reflexes of 
the lower limbs and a spastic condition of the same. 
Sir William Gowers still holds to his opinion, and in the last edition 
of his authoritative work, Diseases of the Nervous System, in reference to 
those cases in which the deep reflexes are lost, says the following: “ A 
few cases have been met with in which a transverse lesion of the dorsal 
cord, especially a transverse concussion-myelitis, without post-mortem 
signs of lumbar inflammation, has caused a persistent loss of reflex 
action in the legs. In these cases, however, there has been not only 
loss of the muscle-reflex action, but also of the skin reflexes—a clear 
indication of an exceptional condition. Moreover, there has also been 
loss of faradic irritability and rapid muscular wasting—in short, there 
have been all the indications of a lumbar myelitis. The most probable 
explanation of these cases is that the descending degeneration of the 
pyramidal tract has been more than usually irritative in nature, so as 
practically to amount to a parenchymatous inflammation, and that this 
has invaded the motor structures as a nutritional change sufficient to 
abolish their function without, however, destroying their form. In one 
such case it was noted that the lumbar nerve-elements were extremely 
granular in aspect.” 1 
Byrom Bramwell, in the latest edition of his able work, Diseases of 
1 From a letter to the author dated April 1,1898 : 
“ Yes, I hold the same views regarding the effectof a transverse lesion on the knee-jerk—that 
there is no interference of a lasting character unless true inflammation descends the cord or 
occurs independently in the lumbar enlargement. It is important to remember that it may 
spread down the cord in the posterior region alone, and this may arrest the reflex action. The 
cases are distinct in which evidence of the occurrence on such secondary inflammation has co- 
incided with loss of the knee-jerk, while the experimental and other evidence that a simple 
transverse lesion does not cause such loss is quite conclusive. Indeed, a single case of the loss 
shows that, when met with, it must be due to some additional element beyond those of a simple 
transverse lesion.—W. R. Gowers.” 12 
the Spinal Cord, says: “ Within the last few years it has been stated 
that if you completely remove the control of the brain from the reflex 
centres in the spinal cord, instead of producing exaggeration you pro- 
duce abolition of the deep reflexes.” Dr. Charlton Bastian, who 
advances this view, states that when the spinal cord is crushed and 
completely destroyed in the dorsal region (when, for example, its whole 
thickness is absolutely destroyed by a transverse myelitis) the deep 
reflexes in the lower limbs are completely and permanently abolished, 
even though the lumbar enlargement is not directly implicated by the 
lesion. He maintains that when the lesion completely severs the con- 
nection between the brain and the lower part of the spinal cord the 
reflexes in the portions of the cord below the lesion are permanently 
abolished. With Dr. Hughlings-Jackson he believes that the cere- 
bellum exerts an enforcing influence upon the lower (spinal) reflexes. 
Dr. Hughlings-Jackson and other good observers have published cases 
in support of this view, which was generally accepted only a few years 
ago. 
“ But whether Dr. Charlton Bastian’s view is correct or not—and I 
must say that, speaking for myself, I have not met with any cases 
which support it—lesions which involve the crossed pyramidal tracts 
are, in the vast majority of cases, attended with exaggeration of the 
deep reflexes.” 
Holt and Herter (The American Journal of the Medical Sciences, 
April, 1895), in referring to a case in which they considered, that there 
was a total transverse lesion of the cord, say: “ The analgesia below 
the neck, the spasm of the legs, and the heightened reflexes are, of 
course, merely the usual results of a total transverse lesion of the cord 
in the cervical region.” 
It would be useless to go further, for up to the time of Dr. Bastian’s 
later contribution (Medico-Chirurgical Transactions, 1890) it was gener- 
ally held by all authorities that a total transverse lesion in the cervical 
or upper dorsal portion of the spinal cord, unattended with degenera- 
tion of the lumbar enlargement, produced an increase in the deep 
reflexes and spasticity in the legs. Notwithstanding the fact that this 
position was generally held, a review of the cases reported in the litera- 
ture at my command furnishes very meagre evidence in support of this 
hypothesis. 
Dr. Bastian’s views caused a general revision of the whole subject; 
so that at present it would appear that his position is the one accepted 
by a large majority of the observers. 
E. S. Reynolds, “ On the Condition of the Reflexes in Total Trans- 
verse Division of the Spinal Cord ” (Brain, spring, 1895), reviews the 
subject very carefully, and finds that such authorities as Bowlby, Thor- 
burn, Bruns, and many others equally noteworthy, agree in the main 13 
with the views promulgated by Dr. Bastian, and reaches the following 
conclusion : “ From the above observations we must see that after total 
transverse division of the cervical or upper dorsal region of the cord 
there almost alivays (italics mine) results a lasting loss of the deep 
reflexes of the legs, with flaccid paralysis of the muscles, and at the 
same time there may be no marked atrophy of muscles, no great altera- 
tion of the electrical reactions, and no affection of the gray matter in 
the lower dorsal and lumbar regions. No satisfactory explanation of 
the phenomena observed is as yet forthcoming, the cerebellar theory 
put forward by Bastian being perhaps the most feasible. 
“ More observations are yet necessary before this subject, which 
affects one of the most fundamental principles in neurology, can be 
satisfactorily settled.” 
Kocher, in his recent and very able work on Injuries of the Spinal 
Cord, is very positive that the deep reflexes are lost when the cord has 
sustained a total transverse division. 
The symptoms in a case of my own, in which there was a total trans- 
verse destruction of the cord from a fracture of the body of the fifth 
cervical vertebra, reported in the Journal of Nervous and Mental Dis- 
eases, June, 1897, agree with the views of Bastian, Kocher, and others 
mentioned. 
D. Gerhardt (Deutsch. Zeitsch. fur Nervenheilkunde, 1894, Band vi., 
pp. 127-136) reports a case which he thinks opposes the views of Bas- 
tian, and points out the fact that if one case of undoubted transverse 
division of the cord can be furnished in which the reflexes remain, Bas- 
tian’s cerebellar hypothesis will be overthrown. 
Reynolds, who abstracted the report for Brain, thinks the case not 
conclusive, as there may have been, while the reflexes lasted, still a 
small portion of gray matter connecting the cord. 
A case reported by Bruns (Archiv fur Psychiatrice, vol.xxv. pp. 1-72), 
in which there was a total transverse lesion of the spinal cord between 
the last cervical and first dorsal vertebrae, shows beyond question that 
the position taken by Gowers in the quotation given in this article is 
not supported in its entirety by the facts in many cases. In Bruns’ case 
there was not sufficient degeneration in the peripheral nerves or in the 
muscles to account for the absence of the deep reflexes, and the lumbar 
enlargement was, with the exception of the descending degeneration 
of the pyramidal tracts, healthy to the most reliable methods of exami- 
nation. 
I single out the case of Bruns because it is considered the strongest 
proof yet given that a total transverse lesion would always produce 
abolition of the deep reflexes and flaccid paralysis of the lower limbs, 
ivithout there being present degeneration in the lumbar enlargement or 
the peripheral nerves or the muscles. 14 
Many other cases of total transverse lesions show the presence of the 
superficial reflexes and faradic irritability; so the views of Dr. Gowers, 
while correct in some essential parts, must be modified. 
In my case it is to be deplored that a more extensive examination of 
the spinal cord could not be made; but, from the positive nature of the 
symptoms, it is really only necessary to establish the fact of the com- 
plete severance of the cord by the tumor to make the case one of con- 
clusive evidence. 
It should be borne in mind that most of the cases which have been 
presented as proof that a total transverse lesion of the spinal cord 
destroys the deep reflexes in the lower extremity, and produces a flaccid 
paralysis of the same, have been traumatic injuries, in which, at the 
same time that the cord is crushed, there are also produced violent 
circulatory and dynamic changes in that portion of the cord below7 the 
injury. In disease-conditions not dependent on traumatism, irritative 
and other forces are at w7ork which must greatly influence the spinal cord. 
The case I have reported here furnishes conclusive proof that in cer- 
tain conditions the spinal cord may be completely severed and yet the 
deep reflexes remain. 
The process by which the division of the cord is effected and the 
attending changes in the cord below7 the lesion are probably the con- 
ditions which cause the reflexes to be retained or abolished. 
In my case the tumor was one of very slow growth, extending over 
nearly seven years. It did not infiltrate the cord immediately around 
it, and must have destroyed its continuity in such a wray as to produce 
the very least possible irritative and circulatory changes in the lumbar 
enlargement. 
In a case to be reported by me in the future, of a large cerebellar cyst 
which was evacuated with recovery of the patient, I offer some evidence 
which might be used against the cerebellar hypothesis of Bastian. In 
this case the deep reflexes were totally absent, and had been for some 
time; but very soon after the evacuation of the cyst, which had 
destroyed by excavation and pressure much of the cerebellum, the 
reflexes returned and the knee-jerks were considerably exaggerated. 
This return of the knee-jerks occurred long before there w7as any indi- 
cation whatever that the cerebellum wras resuming its functions, and 
when the cerebellum was more nearly completely destroyed than in 
any other case heretofore reported in which there was recover}7. The 
destruction of the cerebellum and the attending increase of the knee- 
jerk coincide very closely with the results obtained by Ferrier after 
the removal of the cerebellum in monkeys. 
It should also be remembered that cases have been reported in which 
the lesion did not produce a total transverse division, and yet the deep 
reflexes were lost. 15 
Without attempting to account for the loss of the deep reflexes in 
lesions of the basic ganglia and in the exceptional cases of lesions of 
the pre-frontal lobes and of the cerebellum, beyond attributing the loss 
to the general irritative and inhibitory influences which the lesions 
exert over the entire cerebro-spinal system, I have the following con- 
clusions to offer: 
That the cord may be intact and lumbar degeneration, or degenera- 
tion in the peripheral nerves or muscles, may be absent, and yet there 
may be absence of the deep reflexes. 
That the cerebellum in the human subject may be largely destroyed 
and the remaining portion functionally inactive, and there may still be 
an increase in the deep reflexes in the lower limbs. 
That the spinal cord may be totally severed in the upper dorsal 
region and still increase in the deep reflexes, and spasticity in the lower 
limbs may be present. 
That the rigidity of the paralyzed limbs is not due, as Dr. Bastian 
terms it, to the “ unrestrained,” or, as Dr. Hughlings-Jackson terms it, 
“ unantagonized,” influence of the cerebellum passing downward to the 
spinal cord. 
That the spinal cord is the centre for the deep reflexes, and has also 
within itself the power of producing spasticity of the muscles. 
(I am not prepared to say, however, that the spinal cord has not the 
power, when conditions are favorable—for the transference of power in 
other parts of the body and brain is well known—of centering in itself 
functions which are normally distributed throughout the entire cerebro- 
spinal system.) 
From the foregoing evidence it would appear that the deep reflexes 
are dependent on some power in the spinal cord which severe “ shock,” 
such as injuries and violent disease processes, destroy, not for a few 
days only, but in many cases for all time; and that the loss of the 
deep reflexes is produced by other causes than the total division of the 
cord cutting off the normal influences of the cerebellum or cerebrum, 
or both. 
II. 
A STUDY OF A PECULIAR FORM OF DEGENERATION 
(DEGENERATIO MICANS) FOUND IN THIS CASE. 
EXAMINATION OF UNSTAINED SECTIONS IN WATER AND IN 
SALTY GLYCERIN. 
Around the slit-like cavity in the centre of the tumor mass one can 
in unstained sections, even with the naked eye, see that there exists a 
hand of variable thickness somewhat more opaque than the rest of the 
tumor. 16 
With a magnification of 10 diameters (Plate III. Fig. 2) it is easy to 
make out that this opacity is due to the presence in the tissues in this area 
of minute, irregular particles and rods that look dark on a much lighter- 
colored background, the latter resembling in appearance the main mass 
of the tumor. With this magnification it is evident that (in the sec- 
tion examined) these dark particles are not evenly distributed about 
the slit-like cavity, but are much more numerous in the tissue on one 
side. Even on this side the area in which the bodies are very numerous 
is not of equal width in all places, being from three to four times as 
broad at one extremity of the slit as at the other, while in the areas in 
which the bodies occur there is considerable variability in the number 
of particles present. There is no marked grouping of the particles nor 
is there any sharp line of demarcation between the tissue in which they 
occur and the main mass of the tumor; at the periphery of the band 
the particles are numerous; they grow rapidly fewer in number until 
the band with dark particles and pale background goes over into the 
main tumor mass, consisting entirely or almost entirely of pale back- 
ground. The background of the band containing the particles is, how- 
ever, on the whole, a little paler than the main mass of the tumor, 
which, again, is not entirely free from the dark particles. Here and 
there single particles can be made out in different parts of it. 
With the magnification of 62 diameters (Zeiss apochr. obj. 16 mm. 
compens. oc. 4) one can make out that the dark particles noticeable at 
10 diameters now look dark yellow or yellowish-brown (Muller’s fluid 
hardening), while the pale background looks almost colorless or has 
at most a feeble straw-colored tint. Some of the yellow masses are 
now recognizable as bloodvessels, cut longitudinally, transversely, and 
obliquely, filled or partly filled with fairly well-preserved blood, the 
individual corpuscles of yellow color being distinctly visible with this 
magnification in some of the vessels. (Plate III. Fig. 3.) By far the ma- 
jority of the particles are evidently not bloodvessels, however; they are, 
on the whole, smaller, more refractive, more solid, and much more irregu- 
lar in size and shape. In size they vary from particles smaller than a 
red corpuscle to masses 0.1 mm. and 0.2 mm. in length and 0.05 mm. 
to 0.02 mm. in breadth. Very frequently the large masses are thin at 
one end and thick at the other. They may be straight or may form 
stiff curves. In large part the variable form is evidently due to the 
fact that long masses running in almost every direction through the 
tissue are met with in the section sometimes cut parallel to the long 
axis, sometimes transversely to this axis, or, again, in every degree of 
obliquity. 
Many of the small and some of the larger masses look to be homo- 
geneous, or nearly so, in this unstained preparation. The majority of 
the larger masses, however, present a moniliform or twisted appearance. 17 
The latter, when it occurs in a long, tapering body, thick at one end 
and thin at the other, strongly suggests a screw in shape. Again, a 
long, curved body, with one delicate extremity and another broad 
extremity, may closely resemble in shape and appearance a brass horn. 
With this magnification, what was described above as pale back- 
ground is seen to be far from homogeneous. It consists of (1) longer 
and shorter interwoven rods, fibres, or spindle-shaped bodies; (2) small, 
rather spherical, but less regular masses, which resemble cell-bodies; 
and (3) a somewhat granular, ground substance that presents the 
appearance of a precipitated albuminous fluid. 
With a magnification of 500 diameters (Zeiss apochr. obj. 2.0 mm., 
compens. oc. 4) the details of the structure become much clearer. 
Leaving the peculiar bodies or particles temporarily out of considera- 
tion, the tissue about the cavity is seen to be made up of a mass of cells 
and fibres embedded in a granular material which evidently represents 
the precipitated particles of an albuminous fluid. The elements com- 
posing the tumor are here enormously swollen and loosened, as though 
they had been exposed to the action of some macerating or dissociating 
fluid. The study of the unstained elements making up the mass is 
facilitated by mounting in salty glycerin. The main constituents are 
found to be longer and shorter, narrower, and broader, spindle-shaped 
fibres, containing elongated vesicular nuclei, the latter being situated 
usually, though apparently not always, at the broadest part of the fibre, 
and generally eccentrically, sometimes at the very margin of the cell. 
Portions of two such cells are illustrated in Plate III., Fig. 4, a. 
These fibres, then, are in reality the protoplasm of much elongated 
bipolar cells. A number of structures looking like round-cells, each 
with a small, round nucleus placed eccentrically, are met with in the 
section. These, in all probability, are correctly considered to be trans- 
verse sections of the elongated cells or fibres, the sections happening to 
pass through a part of the cell in which the nucleus is situated. There 
are an enormous number of small round, oval, or polyhedral bodies, which 
vary in size from minute dots as large as the granules of an eosinoph- 
ilous leucocyte to large masses 20, 40, and more microns across. These 
apparently represent cross sections of fibres of different thicknesses cut 
at various distances from the nucleus. One can scarcely ever follow 
one of the long cells throughout its whole course, for the fibres are so 
interwoven, twisted, and curved that each leaves the section at some 
part of its course. Accordingly, what one sees is a complex feltwork 
of longer and shorter segments of cells, some of them containing nuclei, 
others, of course, being devoid of nuclei on account of the plane of the 
section. It is possible that not all of the fine fibrils are really processes 
of bipolar cells. Some of them may be glia fibres, in Weigert’s use of 
the term, since occasionally several can be seen passing through the 18 
protoplasm of a single spherical cell. That the processes of some of the 
bipolar cells may be almost if not quite as delicate as these glia fibres is, 
however, certain, and there are all transitions in thickness from these 
fine fibrils to the large, clear, enormously swollen fibres. Just how many 
of the fibrils represent Weigert’s glia fibrils, and how many are the pro- 
cesses of delicate bipolar cells, cannot be determined. 
It can be made out with this magnification that the particles or masses 
which were distinguishable, even with a magnification of 10 diameters, 
are in large part situated within the swollen bipolar cells, sometimes 
involving only a portion of the cell-body, in other instances apparently 
representing a transformation of the whole mass of the protoplasm of 
the fibre, at least for a portion of its extent. These particles or masses 
as seen in sections in salty glycerin are for the most part of a yellower 
color than the rest of the tissue (even much yellower than the corpuscles 
in the bloodvessels), and are highly refractive. These masses seldom 
look entirely homogeneous. On the contrary, they seem to consist of 
clumps or strands of a waxy-looking substance. They may be entirely 
separate from one another or partially fused with one another, or a 
number may be entirely fused to form a rather large, almost homoge- 
neous mass of the substance. In the screw-shaped or horn-shaped 
masses the clumps or strands tend to be more isolated at the narrow 
end of the screw or horn, while at the broader end they are usually 
fused (or, to be more objective, at the broad end it is often impossible to 
make out that the substance is made up of smaller masses). The long 
masses often exhibit irregular, varicose enlargements in their course. 
In places the fused masses resemble very closely fused and altered 
blood, but with care no mistake on this point is likely to be made, as 
the blood in the capillaries is less refractive and more granular. Be- 
sides, in some instances the refractive masses occupy only a small por- 
tion of the protoplasm of the bipolar cell or fibre, and accordingly could 
have no relation to blood unless one made the improbable assumption 
of a phagocytosis of extravasated blood by the bipolar cells. 
In many of the long cells cut parallel or nearly parallel to their long 
axis the refractive masses are obviously arranged in the form of par- 
allel rows, as though they had resulted from a transformation of longi- 
tudinal fibres or longitudinal areas in the protoplasm. These longitu- 
dinal, parallel masses may be extremely fine and delicate or may be 
coarse. Not infrequently they are visible in the periphery of the cell, 
sometimes apparently on the surface of it, sometimes scattered through 
a part of its substance. 
The longitudinal masses, as a rule, do not extend far without inter- 
ruption, though occasionally they can be followed for a distance of from 
15 to 20 microns. Usually they are very frequently interrupted, and 
resemble, then, minute beads of refractive substance arranged in rows. 19 
The larger minute masses often look fusiform or spindle-like in shape, 
and when lying on the surface of a tapering cell-process in numbers 
remind one strikingly of the gross morphological appearance presented 
by masses of tigroid in the dendrite stained by the method of Nissl. 
Only in gross form and distribution, however, does the resemblance 
hold, for the spindle-shaped masses are yellow or of a straw tint, and, 
as a rule, they are not granular, but homogeneous and translucent. 
The large masses of the refractive substance are often continuous at 
one extremity, with multiple longitudinal bands or multiple rows of 
droplets or spindles of the substance. Sometimes an oblique cut through 
a large mass of the material reveals only partial fusion, and one sees 
projecting from the cut extremity a number of nodular masses arranged 
in rows projecting from a more solid mass, reminding one in form of 
the nodulated, fleshy stems of a cactus. 
In a large number of instances the refractive masses seem to occupy 
the whole area corresponding to the body of the bipolar cell. Where 
this is not the case the protoplasm not involved is easily distinguishable 
from the rest of the mass, being less yellow and less refractive and some- 
times finely fibrillated or granular. As a rule, however, it is of a 
slightly straw-colored tint, and more refractive than the protoplasm of 
the majority of the swollen bipolar cells, which do not contain the 
peculiar masses. One gets the impression that there are many transi- 
tion stages from the pale feeble refractive protoplasm of the least altered 
bipolar cells to the yellowish, highly refractive protoplasm of the cells 
which contain the peculiar masses described. Again, the color and 
refraction of the peculiar masses themselves vary a great deal; whereas, 
the larger masses are very yellow’ and highly refractive (Muller’s fluid, 
section unstained, examined in salty glycerin). The minute strings, 
droplets, or spindles of the substance are often only feebly straw-colored 
and of lower refractive power. Large, swollen, refractive bipolar cells 
occur which are entirely devoid of the peculiar substance; others con- 
tain only a few strands at the periphery or in the substance of the cell 
protoplasm. 
In some of the cells the periphery of the small refractive masses 
appears to be fragmented and flake-like in appearance. 
Outside of the large bipolar cells there can be made out large num- 
bers of minute fusiform or moniliform masses of the substance precisely 
similar in appearance to those which tend to run parallel to one another 
inside the cells. It is extremely difficult to say whether these represent 
totally transformed bodies of delicate bipolar cells or are altered glia 
fibres in the sense of Weigert. One could easily be tempted to form 
the hypothesis that the masses both inside and outside the bipolar cells 
are altered Weigert’s glia fibres ; but while a certain number of bipolar 
cells contain within their protoplasm what" appear to be unaltered glia 20 
fibres, in the majority it is impossible to see such fibres, and in many 
it is difficult to discern even any trace of a longitudinal fibrillation. 
Again, in many of the large, spindle-shaped cells, as has been men- 
tioned above, the whole mass of protoplasm seems to be replaced by the 
peculiar refractive substance. At present, therefore, a simple objective 
description seems more desirable than the making of hypothetical expla- 
nations. 
In many of the bipolar cells the nucleus has undergone a marked 
swelling, accompanied by extreme vacuolization, but no definite rela- 
tion, at least in the unstained preparation, between the nucleus and the 
peculiar masses has thus far been established. 
Action of Reagents upon the Peculiar Masses in Celloidin 
Sections, (a) Water has no effect on the bodies in a celloidin section. 
(6) Nitric acid in concentrated solutions rapidly removes the yellow 
color from the masses as well as from the rest of the tissue, but in other 
respects they are very resistant to the prolonged action of the cold acid. 
With hot nitric acid they disappear, but not before the rest of the new 
growth. 
(c) Caustic Potash. A 10 per cent, solution of KOH swells the masses 
very slightly, if at all, and the form is still easily recognizable after 
action for several hours. They are even very resistant to hot KOH 
(10 per cent.), appearing longer than the other elements of the tumor. 
Finally, however, they swell up, dissolve, and disappear from view. 
This test easily distinguishes them from myelin. In the first place the 
resistance of myelin sheaths to the potash is much greater than that of 
these masses, and, further, while the myelin sheath becomes more refrac- 
tive in the hot KOH, the peculiar bodies in the bipolar cells lose their 
refractiveness and become dull looking. 
Many of the cells of the tumor contain in their protoplasm large 
numbers of minute granules a little smaller than eosinophilic granules. 
These are very resistant to hot KOH, and, on account of their persist- 
ence and high refractive power, are striking features on high-power 
examination of a specimen thus treated. 
(d) LugoVs solution. The masses stain brownish or reddish-brown. 
On addition of 25 per cent, sulphuric acid the color is not much 
changed, although the bodies assume, perhaps, a somewhat deeper 
reddish-brown tint. 
Study of Stained Preparations. Weigert-Pal staining (Plate III. 
Fig. 4). In sections stained by this method a large proportion of the very 
refractive masses stain black. If the differentiation be pushed the tint 
may become brownish or even yellowish-brown in some of the finer par- 
ticles of the substance ; but with ordinary staining all the particles, both 
large and small, stain jet black. The tint assumed is strikingly dif- 
ferent from that presented by the stained myelin of medullated nerve- 21 
fibres, the latter, as is well known, staining of a distinctly bluish-black 
tint. Even with very low magnification (10 diameters), the black, 
stained masses stand out prominently on a yellowish background 
formed by the rest of the tumor. 
With higher magnification (62 diameters) the same curiously shaped 
masses (in the form of rods, flasks, screws, and horns) can be made out, 
as have been described in the unstained preparation. Here all are 
stained black. In some instances at the edges of the black masses can 
be seen yellowish refractive borders, corresponding to the protoplasm of 
the fibres, in which the black masses are situated. Here and there 
quite large, refractive yellow masses, with a tendency toward a spherical 
or oval shape in the sections can be seen, containing scattered through 
them larger and smaller amounts of the substance, which stains black. 
The blood in the bloodvessels and the extravasated blood in the tissues 
stain very differently from the refractive substance under consideration, 
and there is no trouble in differentiating between the two. For the 
details of the structure, however, higher powers and eventually oil im- 
mersion lenses must be employed. A magnification of 500 diameters 
(comp. oc. 4, apochr. obj. 2.0 mm.) suffices to make nearly all the rela- 
tions clear. In specimens so studied it is clear that nearly all, if not 
all, of the very refractive yellow substance seen in the unstained prepa- 
ration stains jet black or brownish-black in Weigert-Pal. As before 
described, this substance appears under various conditions; in solid, 
non-analyzable masses (Fig. 4, b), and in particles, droplets (Fig. 4, o), 
rods, and spindles (Fig. 4, g), of varying size and distribution, being 
almost without exception inside the bodies of the bipolar cells. Very 
often the droplets, rods, or spindles are arranged in rows (Fig. 4, c, d, 
e, f, h) upon or near the surface of the cells or within the substance, 
and one gains the impression that the type arrangement is fibrillary, 
the curious relations met with resulting from granular breaking up of 
the fibrils, with lesser or greater fusion of the broken-up parts, the 
fusion occurring in either a longitudinal or a horizontal direction in the 
cells. Occasionally one of the rows of rods, dots, or spindles takes a 
tortuous course inside the straighter process of a bipolar cell, and may 
even form a spiral inside the protoplasm of the latter. Sometimes 
running among the cells of the growth one can see very fine fibrils, 
which in ordinary gray matter would probably be taken to be glia 
fibrils which have undergone partial or total change into the refractile, 
black-staining substance (Fig. 4, n). Some of these fibrils look intensely 
black, as though consisting of a homogeneous black substance; others 
appear to be made up of a row of irregularly sized granules or droplets 
of the black substance, separated from one another by pale, yellowish 
interspaces. Other fibrils nearby may he unaltered, consisting of feebly 
yellow-colored homogeneous protoplasms. In the large wedge-shaped 22 
or conical masses the broad end often stains uniformly black, while at 
the narrow end the black substance resolves itself into more or less par- 
allel rows of black or brownish-black rods, dots, or spindles. Occasion- 
ally the smaller fragments can still be recognized through the whole 
of one of the larger bodies. When a fibre (i. e., process of a long, 
thick, bipolar cell) running lengthwise in the section is cut somewhat 
obliquely through its broad extremity one can see the homogeneous 
mass of black substance projecting from the cut end. (Fig. 4, o.) 
Sometimes it occupies the whole cross-section of the fibre in its broad 
part; at other times it is surrounded by a peripheral area of protoplasm 
which has not been transformed, but which looks more refractive, as a 
rule, than that of adjacent cells. (Fig. 4, e.) Sometimes this proto- 
plasm or unblackened part of the cross-section of the fibre consists of 
two sorts of substance: a peripheral, more refractive substance, and an 
internal, more pale and less refractive substance. In this paler sub- 
stance are often situated smaller fragments of the black substance seen 
in cross-section. Entirely corresponding appearances are met with in 
longitudinal section through cells altered in the same way. 
From the study of numerous places in the sections one gets the 
impression that the large spherical or oval bodies met with may repre- 
sent areas in which the protoplasm of the fibres has “ run ” into a large 
mass of the substance, for in these masses one can see black dots, rods, 
or spindles, arranged in various ways (Fig. 4, k); sometimes in spirals, 
sometimes in more or less concentric rows, sometimes irregularly dis- 
tributed. It is possible that the appearance may occasionally be due 
to the coiling up of the whole fibre, with apparent fusion of the coiled- 
up parts; more often it looks as though the substance of the fibre had 
“ run” to collect in the larger mass, and so to form a huge, varicose 
swelling in the course of the fibre, or perhaps, by breaking through the 
cell limits, to form a large droplet outside of the cell. 
That the long fibres are in reality portions of the protoplasm, or actual 
bipolar cells, there can be no doubt. Now and then near the cavity in 
the section, possibly on account of the maceration which the tissue has 
undergone, single, isolated cells can be studied, showing the presence 
not only of a nucleus, but also of long, conical processes at each pole. 
(Fig. 4, a.) 
In some of the cells, both in longitudinal and in cross-section, evidences 
of alterations other than fusion can be made out in the black particles. 
Thus in some cells there appears to be a breaking up into extremely 
minute, brownish-black granules, which become more or less irregularly 
distributed. In other cases changes in a single rod or spindle of the 
black substance can be easily observed. Peripheral layers or flakes sep- 
arate off from the main mass, appearing as bluish-gray or slate-colored, 
swollen, flake-like masses adjacent to a central core of very solid or 23 
homogeneous black substance. Again, such a central core may be 
entirely wanting, and a few flakes can be seen in a much paler back- 
ground, the latter being formed in part of the black substance which 
has gone over into solution. In other places the individual rods or 
spindles are swollen, and seem to consist of two sorts of substance—one 
staining black or brownish-black in Weigert-Pal, another very pale, 
almost colorless substance, which stands out sharply from the ordinary 
protoplasm of the fibre in which the rod or spindle is contained (Fig. 
4, l, m). 
Occasionally one of the rows of rods, dots, or spindles takes a tortuous 
course inside the straighter process of a bipolar cell, and may even form 
a spiral inside the protoplasm of the latter. 
It is surprising how, in the very finest fibrils studied, the pale sub- 
stance can be made out in addition to the substance which stains black. 
If very fine fibrils be studied (processes of very delicate bipolar cells ?) 
one can sometimes make out a marked swelling, showing not infre- 
quently varicosities in its course; this swollen portion of the fibril is 
made up of what looks like a network of the brownish-black substance, 
arranged in the form of a mantle, the cavity of the mantle and the 
meshes of the network of the mantle itself being filled up with a pale, 
translucent substance. As a rule, however, the appearance is not that 
of a network, but of a minute number of extremely delicate rods, 
spindles, or dots. Even in very fine fibrils or processes of bipolar cells 
these occasionally become separated from one another, but tend to 
remain arranged in longitudinal rows. 
This method of staining with Weigert-Pal permits one to recognize 
the alteration with the greatest ease in parts of the tissue where only 
single cells are affected. It is found that the change is present not 
only about the cavity in the tumor, but here and there in different parts 
of the growth, and, what is still more important, here and there in the 
white and gray matter of the spinal cord outside the tumor, even at 
distances well removed from it. The appearances presented in the latter 
region are precisely similar in individual instances to those met with in 
the tumor itself. 
Weigert-Pal staining; counter-staining with van Gieson’s method. Cer- 
tain details with regard to the fibres, altered and unaltered, can be better 
brought out by this mode of preparation. Thus the protoplasm of the 
bipolar cells stains quite differently in different cells and in different 
parts of the same cell. The majority of the cells stain only feebly or 
not at all; but some of the larger swollen and more refractive cells 
show an affinity for the acid fuchsin staining of a dull red tint. It is 
especially the periphery of the cells which contain the large fused homo- 
geneous masses of the black substance which takes the acid fuchsin 
stain. Occasionally these refractive peripheries may assume a bright 24 
red color, almost like that of collagen stained by the same method. 
Not infrequently between this broad, red periphery and the black mass 
inside the cell a paler area of protoplasm can be seen. 
Whereas the red periphery in some instances appears to be homoge- 
neous in structure, in others it looks as though it were made up of red- 
dish balls or blocks. The contour of the cell may sometimes, accord- 
ingly, instead of being perfectly smooth, look slightly jagged, the 
indentations corresponding to the lines of junction of the constituent 
balls or blocks. 
Staining with Altmann’s aniline fuchsia; differentiation with picric 
acid. The refractive masses which stain black in Weigert-Pal take 
on a red tint with acid fuchsin alone, and hold it much longer than 
the rest of the tissue on differentiation with aqueous solution of picric 
acid. If the differentiation be pushed, however, the masses lose their 
red color; some of the nuclei of the bipolar cells contain masses which 
stain reddish or of a magenta tint in the acid fuchsin. 
Van Gieson’s stain. With this stain the appearances vary a great 
deal with the degree of picric acid differentiation and with the length 
of time which has elapsed after the specimen has been stained. In 
general it may be said that the less the picric acid differentiation the 
more the refractive bodies retain the acid fuchsin, while in much differ- 
entiated specimens they lose their acid fuchsin entirely or almost entirely 
and look bright yellow. In intermediate stages of differentiation one 
sees varying amounts of substances which have assumed the red and 
yellow or yellow-green tint. Again, as time elapses after the prepara- 
tion of the specimens they gradually grow paler in tint, the acid fuchsin 
apparently being gradually extracted, and the bodies becoming even 
more yellow or yellow-green in cast. 
A moderate degree of differentiation serves best for high-power studies, 
since in preparations so made it can be seen that the large fused, homo- 
geneous masses have assumed a yellow tint, while the smaller rods, dots, 
or spindles may be either yellow, yellowish-red, or altogether red. The 
protoplasm of the bipolar cells remains almost colorless or slightly pink 
in the majority of cells. In the larger swollen cells, however, there may 
be a marked affinity for the acid fuchsin or rather for a mixture of the 
acid fuchsin and hsematoxylin, since the cell-bodies stain of a rather 
dull, dirty, purplish-red tint. Fragmentation of this protoplasm into 
masses, balls, and clumps is frequently met with. 
In some of the better preserved cells a distinct fibrillation can be 
made out in this protoplasm, and the fibrils can be seen, in longitudinal 
section, running more or less parallel to the central picric acid mass, 
while in cross-sections they are observed at the periphery of the picric 
acid mass, the cross-sections of the fibrils being stained of a darker tint 
in the acid fuchsin than the rest of the protoplasm. In some cases the 25 
central picric acid mass is surrounded by a layer of smaller picric acid 
bodies, and outside of these, again, can be seen acid fuchsin bodies. In 
some places these all seem to be within the protoplasm of a single cell, 
but in other places the various bodies seem isolated and not contained 
in a common protoplasm. Whether the latter appearance is due to 
fragmentation of a common protoplasm or to the simultaneous involve- 
ment of a bundle of processes of one or of a number of fine bipolar 
cells, it is difficult to say. One gets the impression that bundles of 
fibrils or of processes of bipolar cells are common; sometimes in the 
centre of such a bundle there is a large fibre, the peripheral elements 
being much smaller and more delicate. These are met with (1) entirely 
unaltered; (2) with alteration in the central fibre; (3) with alteration 
of the central and a portion of the peripheral fibres, or (4) with altera- 
tion of all the fibres. It is not impossible that these various fibres, 
arranged in the form of a bundle, converge to form the protoplasm of a 
single large fibre. If this really occurs one would conclude that some 
of the bipolar cells split up at their poles into a number of finer sub- 
divisions—an idea which agrees well with the illustrations of the find- 
ings of investigators who have worked with maceration methods. 
Staining with iron hcematoxylin. The substance which stains black in 
Weigert-Pal preparations and which stains bright yellow in picric acid 
stains of a slate-gray or blackish-gray color in iron hsematoxylin. This 
method is of particular advantage for the study of the nuclein in the 
bipolar cells. 
In looking in the bibliography for references to similar degeneration 
one is surprised to find how few articles there are which deal with 
degeneration of glia. Weigert, in his exhaustive monograph, makes 
the following reference to “ granular breaking up ” : 
“ The fibres are quite smooth, without granular consistence and 
without circumscribed swellings and thickenings. This holds, however, 
only for specimens examined perfectly fresh or after careful hardening. 
If, on the other hand, a spinal cord be examined which has been cut 
through when fresh, so that the white substance on the surface has 
become swollen so as to show the cadaveric swelling of the myelin 
masses, one can be perfectly sure that in the fibres he will find granular 
breaking up (or that the fibres will be no longer stainable, etc.). This 
cadaveric breaking up of the fibres was first described by Frohmann, 
although Virchow had earlier made the general statement that the 
neuroglia is destroyed by post-mortem influence. 
“ The granules met with in cadaveric breaking up are at first small, 
lying along the long axis of the fibres ; when the injury is more marked 
the granules are larger, small droplets flow together, and the larger 
drops which have originated in this way lie further apart and irregu- 
larly distributed. Finally, they appear to dissolve; at any rate, in 26 
very bad pieces no staining succeeds any longer. The granules of the 
earlier stages of disintegration also stain with more difficulty than do 
the normal fibres. 
“ The varicose neuroglia fibres (cell processes), which many authors 
(on using Golgi’s method) picture and describe, are probably nothing 
but fibres which have undergone previously the cadaveric alteration. 
“ How the granular breaking up occurs is questionable. In my pre- 
vious communication, in the year 1890,1 called attention to the fact that 
cadaveric swelling of the myelin appears to play a role in this connec- 
tion. At any rate, as I then stated, the white substances are those 
which first show the disintegration. It is conceivable, of course, that 
the cadaverically softened neuroglia fibres are ruptured by the swelling 
of the myelin sheaths.” 
The description of Frohmann referred to by Weigert reads as follows :l 
“ The existence of a finely granular gray-ground substance to the 
gray matter is asserted with especial emphasis. I have not been able 
to perceive any trace of such a finely granular substance in a number 
of very beautiful preparations which I have examined. That it, how- 
ever, is found sometimes in the gray substance there can be no doubt. 
Nevertheless, I do not hold it to be a normal constituent of the neu- 
roglia : On the one hand, here, as in the white substance, confusions 
with fibre cross-sections may slip in, which, where they are closely 
crowded, lend a granular appearance to the tissue, and in thicker sec- 
tions and imperfect illumination are with difficulty distinguishable 
from simple granules ; but, on the other hand, not infrequently the fine 
fibres of the gray substance break up into a number of pale, smooth- 
contoured white or grayish granules. This occurs especially distinctly 
where the tissue has been torn, and there one can see, besides the 
granules, also smaller or larger fragments of fibres which have lost 
their glistening appearance as well as their fine but definite contours. 
“ In the continuity of the section the individual fibres are no longer 
distinctly separable from one another, and with the free granules lend 
to it a dull, grayish appearance. I hold these alterations to be the 
product of beginning decomposition—a post-mortem phenomenon which 
can also frequently be observed in the cortical layer. In favor of this 
view, speaks, it seems to me, the fact that I met with a finely granular 
substance most frequently in the ventral and dorsal horn, where the fibres 
are of great delicacy; while they were not seen at all or were met with 
but rarely in the dorsal commissure and in the substantia gelatinosa, 
where the fibres are coarser. Free spherules and spherules not attribu- 
table to the breaking up of fibres were seen by me only in preparations 
containing myelin; they were clear and glistening, occurred only singly 
1 Frohmann, C.: Untersuchungen ueber die normale und pathologische Anatomie des Rtick- 
enmarks Theil i., Jena (1864) S. 49-50. 27 
and scantily, were not stained by carmine, though the larger ones, at 
least, were colored yellowish with iodine. These must, accordingly, be 
held to be small myelin spherules.” 
The similarity of some of the appearances in the case here described 
to this komige Zerfall of Frohmann and to the descriptions of the process 
by Weigert is obvious. The fragmentation of the fibres, the formation 
of varicosities, etc., are striking features in this resemblance. The de- 
scriptions of both Frohmann and Weigert, are, however, so meagre that 
it seems scarcely possible to decide as to the exact relation of the changes 
in the present case to their komige or kadaverdse Zerfall. 
The description of glia degeneration which approaches most nearly to 
that of the case here studied is that of Rosenthal.1 
Rosenthal had the opportunity of studying a peculiar tumor of the 
spinal cord associated with syringomyelia. The tissue contained an 
enormous number of cells of the ependymal type. In it were also 
typical glia cells and glia fibi*es. He describes in this case in the 
close, newly formed glia tissue a peculiar element, abundant and 
striking in certain parts. This element appears in the form of glis- 
tening structures, which assume the shape of small bulbs or twisted 
sausages, with a thicker and narrower end. The largest of them show 
a tendency to lamellation, and resemble corpora amylacea, but do not 
give the typical reaction. 
The staining reactions are peculiar. Tissue hardened in Zenker’s 
fluid and stained by the method of van Gieson contains the bodies 
stained of a brownish-yellow color, sometimes of almost a greenish tint, 
often with a red contour. In specimens hardened in Muller’s fluid 
they stain bright red by van Gieson’s method—quite differently from 
the haline degeneration of the walls of the vessels. Bv the method of 
Weigert-Pal they stain jet black without blue tone. If the decoloriza- 
tion be pushed they stain no longer black, but look dark brown, although 
they are still sharply marked off from the light brown surrounding 
tissue. Weigert’s glia stain was applied to Zenker preparations and 
stained the structures blue black or bluish-gray, with a blackish-blue 
contour. 
In Weigert’s fibrin stain the masses retain a bright blue color. They 
are most distinct in iron htematoxylin preparations, in which they are 
stained jet black in the dull gray sections. This reaction, however, is 
applicable to Zenker preparations only, not to the Muller’s fluid prepa- 
rations. In the latter the iron hematoxylin gives results similar to those 
with the method of Weigert-Pal. 
Rosenthal describes the structures, which resemble swollen bulbs and 
sausages, as being made up in large part of round lumps and very fine 
1 Rosenthal, W.: Ueber eine eigenthumliche mit Syringomylie complicirte Geschwulst des 
RUckenmarks. Beitriige z. Path. u. path. Anat., Jena, 1898, Band xxiii., S. 111-143. 28 
spherules, which often lie together in rows or groups. In some of the 
preparations he sees a common contour about the darker staining 
spherules, and believes that the spherules represent parts of the struc- 
tures which stain in toto in van Gieson’s preparations. He finds the 
structures accumulated particularly about the larger vessels, wThere, 
together with undoubted glia fibres, they form a thick ring on the 
outer surface of the connective-tissue sheath of the vessel. 
Curiously enough, Rosenthal could make out no relation between these 
structures and the cells. He found, however, that they were present in 
the periphery of the cord outside the tumor, where, after secondary 
degeneration, there had been glia proliferation. Benda, who, as Rosen- 
thal states, called his attention to these structures, had repeatedly 
observed these bodies in gliomata, and thought that they represent 
some pathological new formation or attempt at new’ formation of 
myelin sheaths. 
Rosenthal, however, comes to the conclusion that they are significant 
of a form of degeneration of glia fibres in Weigert’s sense. In favor of 
this view he emphasizes the following points: 
1. The structures are found only where there has been an over-pro- 
duction of glia. 
2. No relation to cell bodies can be made out. 
3. They resemble closely in form and arrangement irregularly swollen 
fibres. 
4. They stain by the same methods and in the same color as glia fibres. 
5. They have been found in two cases in tumor tissue in which, along 
with cells and glia fibres, no other element (nerve-fibres, myelin sheaths, 
connective-tissue) was demonstrable. 
Rosenthal designates the process as “ kolbige degeneration,” and 
thinks that many corpora amylacea belong here, especially those 
which do not yield the iron reaction. He states that he can find no 
reference in the literature to structures resembling this degeneration 
of glia fibres, and suggests that the lack of data is dependent upon 
the meagre knowdedge concerning glia which existed before Weigert’s 
studies. 
A comparison of the description of the case now described w’ith that 
of Rosenthal shows many points of agreement, but some divergence. 
The staining reactions, the gross form relations, and the general distri- 
bution of the bodies are similar in the two cases. In the present case 
a number of details observable with oil immersion lens have been de- 
scribed which finds no counterpart in Rosenthal’s description. A 
fundamental difference, too, is found in the descriptions as regards 
the relation to the bodies of cells, whereas Rosenthal states that no 
relation to cell bodies can be made out; in the tissue from Dr. Hud- 
son’s case a very large proportion of the structures are undoubtedly 29 
within the bodies of bipolar cells. Particles of the substance yielding 
precisely the same staining reactions can be found even within the 
nuclei of some of the bipolar cells. The data are as yet totally insuffi- 
cient for the making of any statement with regard to the existence of 
any constant relation between nuclear function and the appearance of 
the substance in the protoplasm of the cells or fibres. Glia fibres in 
Weigert’s sense appear to have been present in much greater numbers 
in Rosenthal’s tumor than in the case here described. Unfortunately, 
the mode of hardening in the present case precluded the application of 
the glia stain of Weigert. As stated in the description, however, fibres 
probably to be regarded as glia in Weigert’s sense are to be met with, 
and a portion of these seem to have undergone the change. 
The term “ kolbige degeneration,” as applied by Rosenthal, seems 
scarcely suitable, as only a portion of the masses resemble a flask in 
shape, and the alterations in the finer fibres certainly do not fit in with 
any such designation. Since the alteration is always associated with 
swelling and the appearance of refractive masses which have specific 
staining reactions, it is desirable to use some term which will indicate 
these features. An entirely suitable term I have not been able to find. 
As I am convinced, however, that the degeneration in Rosenthal’s case 
and in that of Hudson represent identical processes, I would suggest 
that the change be called at present simply Degeneratio micans, or, if 
preferred, Rosenthal’s degeneration, since heffirst described the specific 
staining reactions. It is possible, however, that the acquisition of fur- 
ther knowledge will later on permit of a more satisfactory nomen- 
clature. 
The difficulty in Rosenthal’s case and in that here described in coming 
to an absolute decision with regard to the nature of the elements met 
with (processes of cells or glia fibres), and the undoubted presence of 
enormous numbers of cells in which fibrils do not appear to exist, is in 
accord with a number of cases which have been reported. In this 
connection, and with especial reference to the occurrence of transitional 
forms between cells with simple protoplasmic processes and cells from 
which glia fibres have become fully differentiated, the paper of Taylor1 
may be consulted with profit. If Taylor is correct in his conclusions it 
would be tempting to assume that some of the longitudinal rows of 
particles of the degeneratio micans in the cells in Dr. Hudson’s case 
correspond to the position of differentiating but not yet differentiated 
fibrils. 
1 Taylor, Edward Wyllys: A Contributian to the Study of Human Neuroglia. Journal of 
Experimental Medicine, Baltimore, 1897, vol. ii. pp. 611-634. 30 
III. 
a. NOTES ON DR. HUDSON’S CASE OF SYRINGOMYELIA. b. REGARDING 
THE LITERATURE AND PATHOLOGY' OF SY'RINGOMY'ELIA. 
The following description is intended to supplement that given by 
Dr. Hudson in extenso in the previous pages. Dr. Hudson has covered 
the general anatomical and pathological points so fully that there remain 
to be added only some details of structure of the tumor. The literature 
is not treated in an exhaustive manner owing to its volume, which is 
so great as to exclude it, but especially because it has recently been 
reviewed in the excellent paper of Hoffmann and the monograph of 
Schlesinger. The references made, therefore, are of general interest, 
or refer to cases in which certain of the unusual features met with in 
this instance have been observed. 
The manner of preservation of the tissue and the several modes of 
staining the sections have been mentioned. In addition good and help- 
ful results have been obtained through the use of Mallory’s phospho- 
molybdate-hsematoxylin. 
It will be possible to give the additional details of structure in general 
and without reference to the independent sections, as Dr. Hudson has 
described the latter so fully. 
The description of the tumor will refer (a) to the cells composing it 
chiefly; (6) to the bloodvessels; (c) to the relation of cells to vessels; 
and (d) to the cavities. 
The cells of the tumor are not uniformly distributed throughout the 
mass; they are collected together in masses and in smaller foci. The 
smaller areas show regularity in that they are collected about small 
bloodvessels. The preservation of the cells is not all that could be 
desired, and yet many details can be made out. First, as to form : The 
cells are oval or more nearly round; sometimes they appear in more 
irregular forms, and may be said to be pyriform. Very little distinc- 
tion is possible between cell protoplasm and nucleus, but I attribute 
this to the staining and, perhaps, mode of preservation. The cells lie 
close together or they are separated by areas of lighter appearance. 
Even where they are closely approximated they are separated by inter- 
cellular substance. This substance is fibrillated ; the fibrils are delicate, 
straight or slightly curved, or wavy and sharply contoured. The lighter 
areas are made up of larger numbers of fibres; these being in predomi- 
nance. They are direct, and come off in groups from the cells, running 
in nearly parallel rows until they are lost in another group of cells. 
Where the cells are distributed more irregularly there is an alternation 
of small foci of cells and areas of fibres, the latter being continued 31 
among the cells. No other kind of intercellular substance is to be 
made out between the cells. 
The relation of the cells and fibres to the bloodvessels to be described 
is interesting. Two appearances which depend upon degrees of the 
same condition are met with. In the first the vessel wall is enclosed by 
a felted mass of the fibres and is removed considerably from the lines of 
cells. Depending upon the direction of the section of the bloodvessel 
the fibres proceed to it in a rectangular (transverse) or in a concentric 
(logitudinal) direction. While the fibres are felted there is no great 
complication in this arrangement. In the second it is the cells and not 
the fibres which enclose the vessel wall. The cells here are at least 
several rows deep, and they form a dense collar about the wall. There 
is, however, usually, perhaps invariably, an inner, if narrow, line of fibrils 
between the cells and the wall, just as, by careful focussing, fibrils can be 
made out between the cells. A very characteristic appearance is obtained 
when the cells are removed to a somewhat greater distance from the 
vessel; under these circumstances the vessel is surrounded by lines of 
fibrils proceeding from the cells, the latter here often presenting a pyri- 
form appearance, and which come together upon or just before the vessel 
wall. There is little doubt that these cells and fibres are continuous. 
The bloodvessels in the tumor are numerous. They are, indeed, much 
more abundant than seems necessary if their functions were nutritive 
only. They make up a not inconsiderable part of the mass. Their 
characters are various; for the most part they present greatly thick- 
ened and hyaline walls, and their lumina are small. Blood-corpuscles 
can be seen in them. Other vessels, with delicate walls and distended 
with corpuscles, are met with in the same sections. Complete oblitera- 
tion or reduction of lumen to a cleft-like opening are encountered in 
the former vessels. 
The thickening of the vessel wall affects vessels of all sizes, excluding 
capillaries. The walls thus altered are quite or nearly homogeneous; 
only rarely do they show distinct fibrillation, and they are refractive 
and hyaline; distinction into muscular and adventitial coats is want- 
ing. Within the hyaline walls there is deposited in many cases a 
granular, bluish sediment in specimens stained in hematoxylin, which 
lies just next the lumen or some distance removed. Its exact nature is 
more or less conjectural; but it resembles the deposit of lime-salts 
sometimes seen in large vessels. It is highly probable that there has 
taken place in these vessels an extensive hyaline and perhaps calca- 
reous degeneration of their walls. A similar degeneration is not met 
with in the bloodvessels of the remnant of the cord or in the pia mater. 
In the diseased vessels one of the striking features is the persistence of 
the vasa vasorum after the thickening has become extreme, and indeed 
after complete occlusion of the lumen. Occasionally, instead of com 32 
plete homogeneity of the walls a nucleus of elongated form is still 
present, and now and again globules resembling colloid droplets are 
made out in the walls. 
The Cavities. Two kinds of cavities can be distinguished : (a) pre- 
formed ; (6) softened. With the softened cavities are to be ranked foci 
of partial or complete necrosis in which the degenerated elements have 
not as yet been removed. Certain secondary changes have taken place 
in some of the softened cavities, chief among which is hemorrhage, 
although there is evidence of a fibrinoid transformation of portions of 
the necrotic tissue, or perhaps only of the coagulation of lymph poured 
into them. Hemorrhages have also occurred in the pre-formed, epithe- 
lial-lined cavities. 
The pre-formed cavities. Their sizes, number, and location have been 
stated already. Their walls are composed of neuroglia in which the 
fibres are very distinct. The fibres form a wide layer just about the 
periphery of the cavities, thus making up its boundary wall. Very 
few cells are present in the felt-work of fibres, although many of these 
are mere prolongations of cells lying some distance removed. This 
appearance can be compared with the fibrillated layer about the central 
canal of the cord. Upon this layer of fibres the high epithelium lies. 
In favorable sections the epithelium investment is almost perfect; in 
less favorable ones only fragments remain. It is exceedingly probable 
that in many instances the loss is accidental, and probably took place 
post-mortem. Where there are slight defects in the epithelial mem- 
brane, small bundles of fibres project into the lumen ; this, too, I take 
to be accidental. 
When one examines closely the edges of the cavity it is seen that 
among the fibrils there is contained an irregular globular or coarsely 
granular material of moderately high refraction. This material is prob- 
ably a precipitate, and may have come from degenerated neuroglia in 
this situation. However, other evidences of degeneration are wanting. 
A few red blood-corpuscles have escaped into this tissue. 
As has been described already, hemorrhages have occurred in these 
cavities; in the sections recent blood-coagula are contained in cavities 
surrounded by a complete investment of epithelium. The coagula often 
assume the appearance that suggests congeries of bloodvessels within the 
cavities; there are artefacts due to inclusion of columns of blood-cor- 
puscles within fibrinous channels. Neuroglia cells and fibres may be 
incorporated with these channels when disintegration of the walls of the 
cavities has been caused by or resulted in the hemorrhage. 
The softened cavities. Secondary cavities of this nature undoubtedly 
are present in the sections, but it is not always possible to say positively 
that any large cavity originated in this way. It is possible that certain 
of the large cavities, devoid of epithelial lining, with disintegrating walls 33 
or edges, and containing coagula, blood or fibrin only, may represent 
pre-formed cavities which have suffered disintegration. On the other 
hand, the sections show now and again lighter or pale areas, which, on 
close inspection, may be discovered to be centres of disintegration. 
Where the beginning can be made out it is in masses of fibres rather 
than in groups of cells that the disorganization commences. Red cor- 
puscles become extravasated; faintly refractive globules appear, fibres 
break up, and soon only a formless detritus remains. Another appear- 
ance that suggests the beginning of cavities is seen about certain blood- 
vessels (veins ?) and consists of a pale, granular coagulation, not unlike 
coagulated serum, although the material is the coarser granules. In 
the larger foci remains of bloodvessels with hyaline, thickened walls, 
which appear to have been obliterated, are present. 
From the description given by Dr. Hudson of the gross anatomical 
character of the pathological condition found in the spinal cord, and 
from the microscopical features observed, there can be little doubt that 
the pathological formation must be regarded in the light of a tumor. 
The occurrence of a circumscribed newr growth, displacing and destroy- 
ing the normal tissues of the cord, leading finally to its complete disin- 
tegration, is inconsistent wfith any other conception. 
The microscopical appearances leave no doubt as to the nature of the 
tumor; it is a glioma. In some respects, however, it is peculiar, for it 
is exceedingly vascular—indeed, if the growth were a sarcoma one 
would feel justified, in view of the new formation of bloodvessels which 
has taken place, in speaking of an angiosarcoma; but what is more 
peculiar still is the relation between the cells composing the tumor and 
the bloodvessels. It will be recalled that there is a definite grouping 
of the former about the latter. The cells themselves are of interest 
because of their forms. For the moment we may disregard the high 
epithelium lining the cavities and give attention only to the cells of 
the tumor proper. These are polymorphous; some are round, some 
oval, and others pyriform. The last are much commoner than one 
might suppose from a casual examination, for wherever the elements 
are separated—e. g., softened areas, breaks, oedema—this form of cell 
is very numerous. The cells really are more complex than the term 
employed signifies, for besides nucleus and protoplasm there are pro- 
cesses, one or two in number, which belong to them. The round or 
oval cells are also provided with processes, but whether they come off 
from one pole or both poles, or entirely surround them, is not easy to 
settle. The first appears to be the rule. Between the cells and the 
fibres there is not, unless degeneration is going on, any demonstrable, 
definite intercellular substance. 
The arrangement of the cells about the bloodvessels in circles, so as 
to give rise to appearances suggestive of rosettes, seems not to have 34 
been noticed in syringomyelia. It has, however, been observed in glio- 
mata in other regions. I have described a tumor of the retina in which 
the neuroglia cells grouped themselves in rosettes, not necessarily, how- 
ever, about bloodvessels (Johns Hopkins Hospital Bulletin, 1891). In 
a glioma of the brain (Journal of Nervous and Mental Diseases, May, 
1898) I have described a similar condition to that present in the cord in 
this case. The conceptions expressed in the previous instance regarding 
the nature of the cells may, it seems to me, be applied in the present 
one. In order that they may be entirely clear, I will quote from my 
recent paper the views as expressed by me and which are now held 
regarding the embryology and histogenesis of neuroglia. {Ibid.) 
“ The cells destined to become glia are derived in the embryo from 
the medullary plate, and are at first of the same value as the elements 
which eventually produce nerve-cells. The possibility that cells of 
other value, leucocytes and endothelial cells, may, at a later period, 
become interpolated between the elements of ectodermal origin, and, 
under certain pathological conditions, act as tissue formers, while 
admitted by so good an authority as Ramon y Cajal, have been shown 
to be improbable by the researches of Schaffer and v. Lenhoss6k. 
“ The glia cells—or, to use the term proposed by Fish in this country 
and v. Lenhossek in German, astrocytes—take their origin in the lowest 
and highest vertebrates from the ependyma cells, which are known to 
belong to the supporting cell structures of the cord. In certain low 
forms (myxine, amphioxus) the ependyma cells represent the total of 
the supporting cells, while in the higher forms greater or less numbers 
of astrocytes supply the chief framework of the organ. In mammals, 
and especially in man, astrocytes may, it is considered by some, develop 
from a less highly differentiated cell than ependyma cells, which yield 
them exclusively in the lower forms. It is proposed to call the inter- 
mediate cells (undeveloped forms) astroblasts. 
“ The several different forms of glia cells, those possessing long pro- 
cesses, those provided with short processes (typical astrocytes), and still 
others, in which the filaments come off from one or both poles only 
(brush cells), as well as the true ependyma cells, have all the same 
ultimate origin. Since Weigert’s publication in 1890, and more espe- 
cially since the appearance of his monograph on neuroglia, much atten- 
tion has been given to the relation existing between the fibres and the 
cells in neuroglia. 
“ The pictures given by the Golgi’s silver stain seem to show a close 
union between processes and cell bodies, and to indicate that the former 
are mere protoplasmic elongations of the latter. The method of stain- 
ing introduced by Weigert, and more or less modified by Mallory and 
Beneke, would seem to necessitate a modification of these views, since 
by its employment differentiation between fibres and cells in adult neu- 35 
roglia has been rendered possible. According to Weigert, neuroglia, in 
human beings, consists of cells showing protoplasmic processes only dar- 
ing embryonic life; in the matured or adult condition it is made up of 
a mixture of cells and fibres, in which the latter so greatly predominate 
that they are to be regarded as its chief constituent. ” 
Rosenthal (Ziegler’s Beitrdge, Band xxiii. p. Ill) has described a case 
of syringomyelia associated with tumor formation in which the tumor 
consisted in a number of small cavities, lined with high epithelium, and 
of fibres and cells shown to be neurogliar in nature. Rosenthal regards 
the tumor as having originated from ependyma cells, and proposes to 
call it neuroepithelioma gliomatosum mikrocysticum. The tumor in 
question presented an adenomatous structure, and was, in his opinion, 
developed from the epithelium of the neural canal, which was present 
in several stages of development. This tumor is regarded by Rosenthal 
as unique; it is practically so only as regards the spinal cord, for it 
agrees in certain ways with neuroepithelioma of the retina and some- 
what less with the ependymal-celled glioma of the brain mentioned. 
If we compare the form of cells observed in the glioma of the cord 
with those met with normally in the developing or adult spinal cord, 
there is no difficulty in determining much similarity between the epen- 
dyma cell, especially that observed in the embryonic state, and many 
of the cells composing the tumor. Indeed, I should be disposed to 
regard the tumor as largely made up of such ependymal elements. 
There is significance also in the fact that the tumors of the central 
nervous system, thus far described, in which the ependyma cell has been 
found, show the same tendency to the formation of whorls and to be 
grouped about bloodvessels. 
To come back to the cells lining the pre-formed cavities. These are 
typical adult ependyma cells. Their relations to the other cells of the 
tumor roughly are the relations between the cells lining the central 
canal of the cord and the layer of neuroglia outside. Just as in the 
normal cord this layer is chiefly fibrillar, the fibres representing prolonga- 
tions of the high ependymal epithelium and the modified ependyma cells 
next the canal, which are a part of the supporting framework of the 
organ, so in the tumor the high epithelium lies upon a fibrillar structure 
that originates certainly from the ependvma-like cells of the tumor, into 
which, not improbably, similar fibres derived from the columnar epithe- 
lium may pass. According to this view, the tumor under consideration 
is a glioma of the type described by me under the name of “ ependymal- 
celled glioma.” 
The relation of the cavities to the tumor and the origin of the high 
epithelium lining a part of these can be discussed best after considering 
briefly some of the later views regarding the pathology of syringo- 
myelia. 36 
The oldest and most useful division of cavity formation in the spinal 
cord is into hydromyelia and syringomyelia. There is, however, not 
entire unanimity of opinion as to the limits of each. One of the older 
conceptions regards syringomyelia as a condition in which vertical cavi- 
ties, either entirely independent of or only accidentally united with the 
central canal, exist in the spinal cord. The pathogenesis of this cavity 
formation can be various. In the majority of instances it results from 
the disintegration of a primary new growth of neuroglia. It has been 
customary to speak of this new growth as gliosis or gliomatosis, depend- 
ing upon the extent of the formation and its resemblance to a tumor. 
Saxer ( Ueber Syringomylie: Zusammenfassendes Referat, Cent. f. Patli- 
ologie, 1898, Nos. 1 and 2); Muira (Ziegler’s Beitrdge, 1892, Band xii. 
p. 118) has attempted a distinction between instances of true tumor 
formation (glioma) and mere increase in neuroglia. He thinks the 
cases differ anatomically as well as clinically. However, instances are 
recorded, as in the case here reported, in which, in the same spinal cord, 
besides the existence of definite tumors, less circumscribed growths of 
glia had taken place, and in these cavities had developed. Other cases 
of this kind have been reported by Marchand and Reisinger, Daxen- 
berger and Rosenblath. (See Saxer, op. cit., p. 16.) 
Miura would drop the term “ gliomatosis” entirely; the distinctions 
should be made between glioma, as a circumscribed tumor, and diffuse 
glioma when the growth is more diffuse or general. The literature con- 
tains relatively few cases of cavity formation in the spinal cord which 
have originated in glioma used in this sense. The great majority of 
instances have arisen in more diffuse growths—“ gliosis” or “ central 
gliosis ” of Hoffmann. 
The new formation of glia occupies regularly the central part of 
the cord in the neighborhood of the central canal, and generally its 
location is behind (dorsal) the canal. This regularity of location is one 
of the main points of difference as compared with the tumor growths. 
The new growth takes place in a longitudinal direction, and, in most 
cases, it is tunnelled by a tubular cavity. The cavity, however, may 
vary greatly as regards size and form. All transitions have been noted 
between such a condition as described and instances in which, through- 
out the whole length of the cavity, very little or no increase in glia can 
be demonstrated. The usual conception regards the new growth of 
neuroglia as the primary condition and the cavity as representing a 
degenerative process. 
It is desirable to consider the cause of the primary new growth of 
neuroglia about the central canal which this theory assumes, and to 
inquire whether there is any causal relation between the central canal 
or the central ependyma fibres and the central gliosis. The best answers 
to these questions which can be given at this time are the views expressed 37 
in the monographs of Hoffmann1 and Schlesinger.2 These authors con- 
sider the primary condition to be some congenital anomaly which results 
in the persistence of nests of embryonal tissue behind the central canal, 
perhaps in the line of closure. Hoffmann regards as of equal impor- 
tance for the etiology the existence of multiple spinal canals or their 
“ anlage,” and the occurrence of large cells, the equivalents of epen- 
dyma cells in the posterior commissure in the neighborhood of the 
posterior longitudinal fissure. According to Hoffmann, the new growth 
of glia arises from the central canal epithelium or its equivalent in the 
line of closure or from the “ subepithelial epithelioid ” cells. Hoffmann 
regards, therefore, as the principal factors in the production of the path- 
ological condition to which the name syringomyelia is applied, (1) some 
congenital anomaly of development in the cord, and (2) activity of the 
epithelium of the central canal. 
Schlesinger agrees in the main with Hoffmann’s views. He conceives, 
in addition, that anomalies in the vascular supply are largely responsible 
for the formation of cavities. He says: ‘ * Disease of the vessels plays a 
role co-ordinate with the central gliosis.” Schlesinger also discusses the 
frequency of the occurrence of a lining, partial or complete, of high 
(central canal) epthelium in the cavites, and he concludes that this is a 
much more common phenomenon than has been believed heretofore. 
He would place hydromyelia, with its completely epithelial-lined canal, 
and syringomyelia, with only, at times, a neuroglia or connective-tissue 
lining to its cavities, at the two ends of a single anatomical state. In- 
deed, he goes even further and described cavity formation in previously 
normal cords as syringomyelia, the cavities in these cases being the 
result of vascular disease; and he believes that tubular hemorrhages 
may give rise to a similar condition. 
The views of Hoffmann and Schlesinger have emphasized the part 
played by the central canal epithelium in the production of the central 
gliosis. Saxer (op. cit., p. 52) also attributes a large role to this struc- 
ture. His views may be summarized as follows: There are positive as 
well as theoretical grounds for considering the central canal epithelium 
as capable of taking part in the production of pathological glia forma- 
tion. The process, however, does not proceed in the manner described 
by Hoffmann, but it conforms to the processes met with in the embry- 
onic spinal cord. Moreover, Saxer thinks that in the multiplication of 
the epithelium of the central canal and its equivalent there is produced 
not only neuroglia, but structures of typical epithelial character, such 
as that lining clefts and cavities and the foci of epithelial cells in the 
posterior line of closure. Through this new growth of epithelium there 
1 Deutsche Zeitschr. f. Nervenheilkunde, vol. iii., 1893. 
2 Die Syringomyelie, Vienna, 1895. 38 
may ultimately develop supernumerary or duplicate central canals. 
Saxer does not think it proven that the central canal epithelium is the 
essential source of the new tissue, or that, as a rule, the new growth 
starts from this structure. 
Reference has been made to the view entertained by Schlesinger that 
cavity formation leading to syringomyelia may occur without increase in 
neuroglia. Muller and Meder (Zeitschr. f. klin. Med., Bd. xxviii.) have 
encountered a similar condition. In both profound disease of the vessels 
existed. The irregular and slightly increased amount of neuroglia found 
by Muller and Meder is considered by them to be of secondary origin. 
Weigert (Beitrage z. Kenntniss d. Normalen Menschlichen Neuroglia, 
1895) has brought forward similar cases, and from them he draws the 
following interesting conclusions (op. cit., p. 156): “Ignorance of the 
extent of the thickening of the glia about the central canal has given 
rise to mistakes. It has led to the fable of the softened central gliosis 
in syringomyelia. The confusion has been increased by confounding 
gliosis with glioma. The conception of the ‘ softened central gliosis * 
involves the idea of an increase in typical, fibrillar neuroglia with soft- 
ening ; but the neuroglia is normally much increased about the central 
canal, and, indeed, as compared with this form it may be actually 
diminished in syringomyelia and hydromyelia. On the other hand, 
one finds extensive glioses, as, for example, in multiple sclerosis; but 
these do not soften, and, indeed, it has never been proven that true 
gliosis does soften. In a word, the conception of syringomyelia as 
softened central gliosis has not the shadow of a probability to support it.” 
Saxer (Zeigler’s Beitrage, Band xx.) has attempted to answer this 
criticism, and he has drawn attention to the fact (1) that the process of 
softening has been observed and described step by step, and (2) that 
many cases with great increase in glia, excluded from being of second- 
ary growth, have been encountered. However, it is admitted that 
cases to which Weigert’s description applies do exist. 
These points have been developed with some fulness, inasmuch as the 
case described in this paper covers two of the disputed questions: (1) 
Beginning and advanced softening were observed in the pathological 
glia; and (2) the extent of character of the pathological new growth 
proves it to have been of primary origin. 
The present case is also of theoretical interest in respect to the light 
thrown by it upon the question of the participation of central canal 
epithelium in the production of the new growth of glia. All recent 
writers have dwelt more or less on this feature; apparently only Saxer 
has regarded a return to embryonic conditions in the epithelium as 
probable. Although Saxer takes pains not to overemphasize the theory 
of congenital anomaly in predisposing or leading to the production of 
new tissue, yet his view involves some such condition. Upon theoretical 39 
grounds alone Y. Lenhossek (Der feinere Bau des Nervensystems in Lichte 
Neuester Forschungen, 2 auflage, p. 176-248) has regarded abnormali- 
ties in development as playing a large part; and Hoffmann believed he 
had found the misplaced elements whieh in development gave rise to 
the new tissue. If the views expressed in this paper are correct, the 
new tissue, in the case here given, is made up of cells having the value 
of central canal epithelium, but existing in an embryonic rather than 
in a fully developed state. Further, the arrangement of the cells is 
highly suggestive of that of the embryonic ependyma, the bloodvessels 
in this instance supplying the place of the central canal of the cord. 
Without laying too much stress on the probability of the view expressed 
by Saxer, it is yet worth while to mention that as conceived by him, 
along with the products of embryonal ependyma, those of the developed 
tissue are afforded by the high epithelium lining many of the cavities. 
Finally, it may be well to add just a word on the mooted question as 
to whether the new tissue in this case is to be regarded as a tumor. 
The lines between simple new growth of tissue and that growth that 
constitutes for us “tumor” is not sharply drawn. We regard, how- 
ever, an independent formation of tissue, not arising merely as a protec- 
tive or supporting structure, or in response to the action of some known 
irritant (actinomyces for example), as a tumor. Moreover, tumors often 
differ from simple tissue growth because they occur and persist in a 
state of development foreign to the tissues after a certain period (sar- 
coma). While granulation tissue at a certain stage may resemble 
sarcoma, it will sooner or later cease to do so, because it tends to 
change progressively into adult tissue. Fixity in an abnormal state of 
development is one differential point. Apparently this is the condition 
in the process described in the spinal cord. There has been a great 
increase in neuroglia; but the cells are not the type of adult, devel- 
oped, neuroglia cells, nor are a part embryonic and another part cer- 
tainly adult cells, if we exclude the high epithelium lining the cavities. 
The tumor cells proper are all embryonic and of the type of early epen- 
dymal cells. The circumscription of the growth, its evidently primary 
character, the state of development of its composing cells, would in 
other situations, at least, if the same kind of conditions could be met 
with, lead to the conclusion of tumor formation. M. Allen Starr1 has 
recently reviewed the literature concerning this consideration as well as 
some other questions concerning syringomyelia. 
1 Journal of Nervous and Mental Disease, January, 1897. 40 
DESCRIPTION OF PLATES. 
PLATE I. 
The drawing shows the position within the spinal cord of the new growth of glia and its 
extent; it also indicates the levels of the several sections and their chief characteristics. The 
letters e.c. indicate the epithelial cavities, t. is the tumor. 
Fig. 1 corresponds to Plate I., Section 1. Central growth of glia developed within the gray 
matter of the cord, chiefly within the posterior horns, which are entirely destroyed. Anterior 
horns partially preserved, a and b. Epithelial-lined cavities. Weigert, Pal, and van Gieson’s 
method. Magnified 6 times. 
Fig. 2 corresponds to Plate I., Section 4. The new growth is somewhat more extensive, and, 
while occupying the entire posterior segment of the cord, it also encroaches upon the anterior 
white matter. The central zone shows rarefaction, ab c. Epithelial-lined cavities. Magnified 
6 times. 
Fig. 3 corresponds with Plate I., Section 5. Except for the large number of epithelial-lined 
cavities, a b c d ef, it is similar to Fig. 2. Magnified 6 times. 
PLATE II. 
PLATE III. 
FiG. 1 corresponds with Plate I., Section 6. Two foci (stems) of tumor developed centrally 
in the cord, but posterior to the anterior commissure. The position of the central canal is 
indicated by c. In the larger focus an epithelial-lined cavity, in the smaller, a degenerative 
one, are shown. Magnified 6 times. 
PLATE IV. 
Fig. 1. Portion of Section 4, magnified 500 times. This drawing is from a softened area into 
which hemorrhage has taken place. It shows fragments of fibres, naked nuclei, and free red 
blood-corpuscles. There are independent cells and fibres, probably due to a degenerative pro- 
cess. Stained after Van Gieson’s method. 
Fig. 2. Portion of Plate I., Section 3, magnified 100 times. Projecting folds, resembling 
Grecian columns, covered with cylindrical epithelium, are shown. The folds are composed 
of rosettes of polyform cells, the similarity to ependyma-cells being evident. Bloodvessels with 
much thickened walls and others with normal walls are present. The rosettes are developed 
about small bloodvessels. Stained in haematoxylin and eosin. 
ADDITIONAL FIGURES. 
Plate III., Fig. 2. Section through half of stem of new growth, showing part of cavity and dis- 
tribution of the larger masses of glistening degeneration (Degeneratio micans). Muller’s fluid, 
hardening, celloidin section unstained, mounted in salty glycerin, magnification of 10 diam- 
eters. The branch of the cavity marked x corresponds to the portion seen under higher power 
in Fig. 3. 
Plate III., Fig. 3. Higher magnification of a part of the same preparation represented in 
Fig. 2. 
c. Bloodvessels. The masses of the glistening substance assume various forms. A large num- 
ber of pale swollen fibrils containing none of the refractive substance are indicated. Near 
the central cavity the coagulated albumin is seen. Magnification of 62 diameters (Zeiss, Corn- 
pens. Oc. 4, Apochr. Obj 16 mm.). 
Plate III., Fig. 4. Series of drawings representing elements selected from a Weigert-Pal prep- 
aration of the tissues pictured in Figs. 2 and 3 ; magnification of 500 diameters (Zeiss Compens. 
Oc. 4, Apochr. Obj. Oil imm. 2 0 mm.). 
a. Parts of two swollen bipolar cells met with in one field of the microscope. These two 
cells do not contain the glistening substance. The nucleus is in both placed excentrically, 
this being obvious in the one cell seen *• in profile.” 
b. Part of long bipolar cell in large part converted into the glistening substance, here stained 
black. At the thick extremity is a drop of protoplasm showing none of the Degeneratio micans. 
c. Part of a swollen bipolar cell showing a black core with peripheral, nearly parallel, rows 
of dots and rods of the black substance. 
d Part of a bipolar cell showing rows of drops and rods, as well as several large drops of the 
blackened substance. A few very pale areas are visible in the protoplasm. 
e. Part of bipolar cell showing central core of substance stained black, which at one end of 
the figure is resolved into a bundle of parallel rows of black dots and rods. There is a rela- 
tively large amount of swollen rather refractive protoplasm peripherally placed, which is free 
from Degeneratio micans. 
/. A small bundle of interrupted black lines of the stained glistening substance embedded 
apparently in a common ground mass of protoplasm, 
g. A complex of sections of fibres of various sizes running in different directions. Fusiform 
masses of the black substance distinctly visible in some of them. 
h. Cross section of a rather large bipolar cell showing transverse sections of the longitudinal 
rows of masses of the blackened Degeneratio micans. 
k. Large refractive mass containing particles of the blackened glistening substance. The 
protoplasm has apparently " run ” into this form. 
l, m. Sections of swollen bipolar cells containing sections of pale masses. Inside the latter 
are seen irregularly distributed droplets of the blackened glistening substance. 
n. Two very fine'fibres, both taken from a part in which all the elements in the tissue are 
swollen. One of the fibres is free from, the other contains, the substance which stains black 
in Weigert-Pal. 
o A rattier fine fibre, showing fragments of the blackened substance. The upper thickened 
end of the fibre is seen in cross sections and on the cut surface is seen a central core, which 
consists of a homogeneous fused mass of the blackened glistening substance. Plate I. 
Mai Brodel, del. 
Heliotype Co.,Boston. Plate II. 
Fiy.l. 
Fig. 2. 
Fig.3. 
Max Brodel, del. 
Heliotype Co.,Boston. Plate III. 
Ficf.l. 
Fig. 2. 
Fig.3. 
Fig. 4, 
1 x 5(X) 
Max Brodel, del. 
Heliotype Co.,Boston. Plate IV. 
Fir). 1. 
1 x 500 
FUf.2. 
1 x 100 
Max Brodel, del. 
Hdiotype Co,Bostim. The Medical News. 
Edited by J. RIDDLE GOFFE, M. D. 
PUBLISHED WEEKLY, AT $4.00 PER ANNUM. 
THE practitioner who would give his patients the benefit of the latest and best knowledge, 
and thus most certainly ensure his own success, cannot afford to neglect the medical 
weeklies. Foremost among them, as well as cheapest, stands The Medical News. Its 
32 quarto pages of reading matter provide space each week to lay before its read- rs the latest 
advances in medicine, surgery, obstetrics, and the various specialties. This vast range of infor- 
mation is gathered from the whole civilized world by the perfect organization of The News. 
Its reputation brings to it the contributions of rr cognized leaders of medical thought and practice. 
It is a marvel of modern times that the immense body of the world’s advancing knowledge in 
medicine can be laid each week upon the table of the subscriber at the comparatively trivial price 
of 84 per annum. The secret of course lies in a circulation embracing the most progressive men 
in the United States and Canada as well as in many foreign countries. 
The American Journal 
of the Medical Sciences. 
Edited by ALFRED STENGEL, M. D. 
PUBLISHED MONTHLY, AT $4.00 PER ANNUM. 
BEING the medium chosen by the best minds of the profession during seventy-nine years for 
the presentation of their elaborate papers, The American Journal has well earned the 
praise accorded it by an unquestioned authority—“ from this file alone, were all other 
publications of the press for the last fifty years destroyed, it would be possible to reproduce the 
great majoiity of the real contributions of the world to medical science during that period,” 
Original Articles, Reviews and Progress, the three main departments into which the contents of 
The Journal are divided, will be found to possess still greater interest than in the past. The 
brightest talent on both sides of the Atlantic is enlisted in its behalf, and no effort will be spared 
to make The Journal more than ever worthy of its position as the most useful and valuable of 
medical magazines. 
Progressive Medicine. 
A Quarterly Digest of New Methods, Discoveries and Improvements in 
the Medical and Surgical Sciences. 
Edited by HOBART AMORY HARE, M. D. 
44 f PROGRESSIVE MEDICINE” will be modelled upon an entirely novel plan and will 
present in an original, narrative form a clear statement of the practical advance made in 
every department of medicine and surgery during the year, each specialty being dealt 
with by a single authority whose reputation gives ample assurance of accuracy and completeness. 
The personal impress of the contributor will be upon each subject; his thoroughly digested narrative 
will present its data in due and instructive connection, enchaining the interest, economizing time, yet 
giving ample and detailed consideration to all matters which will prove of real value in actual 
practice. 
Four volumes will be issued each year, the first appearing on March 1st, 1899, and the others 
following at intervals of three months. Each will contain from four to five hundred octavo pages, 
abundantly illustrated and handsomely bound in cloth. 
Expecting appreciation of the self-evident utility of such a work to all practitioners, the pub- 
lishers confidently anticipate a circulation so general as to justify the very moderate subscription 
price of ten dollars for the four volumes. All expenses of delivery will be prepaid. 
$10.00 PER ANNUM. 
THE FOLLOWING FAVORABLE COMBINATION RATES ARE AVAILABLE. 
American Journal of the Medical Sciences $ 4.00 
Medical News 4.00 
Per Annum 
Alone 
$75® 
In Combination 
Progressive Medicine 10.00. 
Medical News Visiting List 1-25. 
$15.00 
$15-75 
$16.00 
Medical News Formulary 1.50, net. 
In all $20.75 lor $16.00 
“Progressive Medicine,” with the “American Journal” or “Medical News,” will be supplied at 
$13.00, or the “American Journal,” “Medical News,” “ Visiting List” and “Formulary” at $8.50. 
LEA BROTHERS & CO., Publishers, 
706, 708 & 710 Sansom St., Philadelphia. 
111 Fifth Aye., New York.