OBSERVATIONS ON THE MARROW OF THE BONE AND 
THE SPLEEN IN A CASE OF LEUK/EMIA. 
BY 
JOHN GU ITER AS, M.D., 
OF PHILADELPHIA. 
Reprinted, from the 
Transactions of the Association of American Physicians, 
1 89 5. OBSERVATIONS ON THE MARROW OF THE BONE 
AND THE SPLEEN IN A CASE OF 
LEUKAEMIA. 
By JOHN GUITKRAS, M.D., 
OF PHILADELPHIA. 
Time 1ms not permitted a complete study of the present case of 
leukaemia. In the following paper attention will be called to some 
observations made upon the structural changes in the marrow and in 
the spleen and to the bearing they may have on the process of cyto- 
genesis. Unfortunately, it has not been possible to study this process 
in the normal organs from the point of view suggested in the paper. 
The case was one of leukiemia, treated in the Philadelphia Hospital 
under the care of Dr. William E. Hughes. The case was diagnosed 
as one of splenic and myelogenic leukaemia. 
The notes of the post-mortem are as follows: 
Isaac L., aged fifty-five years, white, male, native of England, admitted to 
the Philadelphia Hospital, December 11,1894; died February 5,1895. Post- 
mortem made sixteen hours after death. 
Pathological Diagnosis. Anaemia. Hyperplasia of the spleen. Lymphoid 
marrow of the bones. Lymphoid infiltration of the liver and the kidney. 
Body of a very emaciated man. Surface of the body considerably dis- 
colored by freezing, but showing evidences of a slight yellow discoloration. 
Round, superficial ulcers are found over the external malleoli. 
Abdominal Cavity. The liver projects beyond the edge of the thorax 
4 cm. in the right mammary line, 10 cm. in the median line, and 44 cm. 
to the left of the same. The spleen projects 94 cm. in the left mammary 
line, and extends 2£ cm. beyond the median line. The left lobe of the 
liver overlaps the spleen 1 cm. The appendix projects inwardly, and then 
bends down over the psoas muscle into the pelvis. 
Thoracic Cavity. The lungs are free and small, and the pleural cavities are 
empty. The costal pleura is marked by minute spots of dark pigment in the 
intercostal spaces. A few of these are found also in the mediastinal pleura. 
On opening the pericardial sac a dense band of adhesion is found in front. 2 
GUITERAS, 
The right ventricle is covered by a large white fibrous patch. The rest of the 
visceral pericardium shows evidences of myxomatous change. 
Heart. Both ventricles are flabby. The left cavities contain a small 
amount of liquid, pale-rose-colored blood and some grayish coagula. The 
mitral orifice admits two fingers. The right cavities contain a large, soft, 
grayish clot and a small amount of liquid blood. The tricuspid orifice 
admits with difficulty three fingers. The right auricle is somewhat dilated. 
The other cavities are normal in size. The mitral valve is slightly thick- 
ened and presents a few opacities. The other valves are normal. The 
apices of the papillary muscles show some evidences of fibroid change. The 
aorta presents some patches of atheroma near the valves. The myocardium 
presents a brownish color. The orifice of the pulmonary artery measures 
10J cm. in circumference. The pulmonary valves appear somewhat large. 
The aortic orifice measures 9£ cm. in circumference. The heart weighs 
310 grammes. 
The left lung presents extensive anthracosis. The surface of section is other- 
wise normal. The bronchial glands are of normal size, and some of them are 
slightly indurated and surrounded by a thickened capsule. Others are rather 
soft. They are all black in color. The lung weighs 460 grammes. The 
right lung similar to the left. The bronchial mucous membrane of both 
lungs is somewhat swollen, and presents a marble discoloration of a purplish 
and gray color. The organ weighs 680 grammes. 
Spleen. The splenic vein flattened out measures 1.3 cm. in diameter, and 
the artery half a centimetre. The organ is large, dense, and irregularly 
tongue-shaped, the upper extremity being thick and rounded and the lower 
extremity presenting a flattened edge. It presents a large concavity toward 
the hilum. There is a large notch on the posterior border and a smaller one 
on the anterior. The organ measures 25 cm. in length, 14! in width, and 
10 in thickness. At the upper end there is some thickening of the capsule. 
The latter presents a pale-magenta color. The surface of the section is 
smooth, and shows the color of cherry wood. There is no prominence of the 
Malpighian bodies. Toward the cortex the structure is slightly translucent, 
and very minute areas of a darker color and slightly translucent are seen 
throughout the substance. There is no prominence of the trabecula. The 
organ weighs 1810 grammes. 
The suprarenal bodies are soft and dark in color. The connective tissue 
around them shows evidences of mucoid change. 
The left kidney is rather soft, the capsule is easily removed, and the sur- 
face is smooth and grayish-brown in color. The surface of section shows a 
marked contrast between the pyramids and the cortex. The former are of a 
rose-red color, and the latter is gray. The structure of the latter is not clear. 
The organ weighs 148 grammes. The right kidney presents a similar appear- 
ance to the left, and weighs 148 grammes. They both show some reduction 
of the cortex. The color of the latter may be compared with that of dark 
coffee and milk. MARROW OF BONE AND SPLEEN IN LEUKAEMIA. 
3 
The pancreas is somewhat firm, and presents a pale yellowish-gray color. 
It is otherwise normal. 
The stomach is rather contracted. The mucous membrane is smooth and 
presents a pale-rose color. 
The duodenum is normal. The common duct is patulous. The bile is 
fluid, and of a pale-yellow color. 
The liver is rather soft, and the capsule presents a few patches of thicken- 
ing. The color is of a rather pale brown, presenting in some portions minute 
mottlings of a yellowish-gray color. The organ measures transversely 25 cm. 
The right lobe measures antero-posteriorly 22 cm., and the left lobe 19. 
The right lobe is 9 cm. thick. The surface of section is smooth, brown, and 
marked by minute grayish areas in the periphery of some of the acini. The 
organ weighs 1900 grammes. 
The mesentery contains a small amount of fat showing mucoid change. 
The mesenteric glands are very slightly enlarged, of normal consistency, and 
some of them on section present a marbled appearance of white and rose 
color. The inguinal glands are similarly affected. No change is found in 
other lymphatic glands. 
Interline*. The valvulte conniventes are prominent and cedematous. In 
some places they are stained with bile, and in others they show evidences of 
congestion. The large intestine is normal. 
The abdominal aorta is large and presents a few patches of atheroma, 
especially in the lower portion. It contains puriform clots. The thoracic 
duct presents a normal caliber and rather thin walls. 
The testicle is rather soft, and presents striae of a yellowish-brown and rose- 
red color, 
The bladder is normal. 
Bones. The sternum presents a brownish-purple and pulpified marrow. 
The costal cartilages present on section a marked translucent brownish 
central area. The medulla of the femur is rather soft and dark purple in 
color. 
The sections of the marrow and the spleen that have received care- 
ful attention in this case were cut from small pieces that were placed 
for two hours in Midler’s fluid, then washed, fixed in Fleming’s, 
hardened in alcohol, and cut in celloidin. The methods of staining 
were safranin, Benda’s haematoxylin stain, which was used for tissue 
that had been hardened in Midler’s fluid, Biondi’s fluid, and a stain 
of safranin, orange G, and methyl-blue. 
Anyone undertaking the study of these two organs recognizes at 
once the difficulty of the problem before him. The variety of opinions 
expressed as to the nature of the cellular elements is rendered more 
confusing by the variety of methods that are suggested for their study. 4 
QUITE RAS, 
No attempt will be made in the present paper to enumerate and dis- 
cuss these opinions. In the bone-marrow especially we are confronted 
with cells that are diplasmatic to a marked degree. We have here in 
the loose reticulum, which has a tendency to arrange itself in the form 
of tubes, cells of varying size. Some of them contain fat, others 
haemoglobin, others eosinophile granules, others present an osseous 
protoplasm, and others the colorless protoplasm of the amoebocytes. 
A rather ineffectual attempt to dissociate these cells is based upon 
peculiarities of the nucleus, even before the protoplasm becomes char- 
acterized by the different forms of metaplasma. The manner of 
division of the nucleus is insisted upon as an important feature in 
the differentiation of these cells. It is very probable, however, that 
different forms of nuclear division may be found in cells that are not 
fundamentally distinct from one another. 
In the present case the lymphoid reticulum of the bone-marrow is 
found slightly thickened, and spaces are crowded with a variety of 
cells. These may be seen in Figs. 1, 2, and 3, accompanying the 
present paper. Attention is especially called to Figs. 1 and 2, where 
different forms of nuclear division and a variety of cells are shown. 
Nowhere in the marrow has any evidence been found of the formation 
of red blood-cells. There are, undoubtedly, numbers of cells that pre- 
sent the nuclear features that are said to be characteristic of erythro- 
blasts, but it appears that they become overgrown with protoplasm, or 
may present evidences of nuclear cell-division without being able to form 
red blood-globules. The protoplasm of such cells frequently shows un- 
mistakable evidence of the presence of haemoglobin. In Fig. 1 a widely 
dilated endothelial channel is shown filled with such cells. Their nuclei 
present the indirect forms of segmentation of the nucleus that have been 
described by Arnold. There are evidently no karyokinetic figures. In 
Fig. 2 a large marrow-cell is seen to the left, presenting a similar poly- 
morphous arrangement of the nucleus. The nucleus of this cell presents 
a small amount of chromatin and a distinct nuclear membrane. To 
the right of this we find an erythroblast which shows swelling of the 
protoplasm and increase in the amount of chromatin and an attempt 
at mitotic division of the same. Such cells are frequently found in 
the marrow in the present case, and undoubtedly find their way into 
the circulation and swell the number of leucocytes and nucleated red 
blood-cells. The next cell to the right is a resting-cell of the same MARROW OF BONE AND SPLEEN IN LEUKEMIA. 
5 
kind as the one to the left, but in which no fragmentation is taking 
place. We have, finally, to the right of this a large fat-cell. In fact, 
in these wide channels of the lymphoid marrow we may find cells in 
all stages of direct nuclear fragmentation and varying forms of chro- 
matolysis, and cells varying in size from the small round erythroblasts 
(that will be described further on) to the large cells shown in Figs. 1 
and 2. With the exception, perhaps, of the true erythroblasts, these 
cells may be considered as interchangeable, and even the erythroblasts 
appear to be subject to such changes as may bring about their con- 
version into the form of cell described as leucoblast, that is, their 
nucleus may become gradually poorer in chromatin, and will assume 
by degrees the loose reticulated appearance characteristic of the larger 
cells. 
In regard to the spaces where these cells are to be found in the 
bone-marrow we note the fact that many of them are not lined with 
endothelial cells. Some of them, on the other hand, present the endo- 
thelial lining of capillaries. These endothelial cells show no evidences 
of nuclear cell-division. They are oval in shape, poor in chromatin, 
and project slightly into the lumen of the vessel. These endothelial 
channels appear to be surrounded by channels that are not lined with 
endothelial cells. It is in these peri-endothelial channels that we 
shall find, later on, in the spleen, a large number of erythroblasts en- 
gaged in the process of red blood-formation. In the marrow of the 
bone, however, they do not gather in these localities, and therefore 
fail to contribute to the formation of red blood-cells. This is true at 
least of the present case of leukaemia. 
I wish to call especial attention to the cell marked with the letter a 
in Fig. 3. The two nuclei represented in this cell are surrounded by 
a transparent zone. The nuclei are homogeneous and stain intensely 
red with safranin. It will be seen later on that bodies similar to these 
nuclei are especially concerned in the formation of red blood-cells. 
Cells similar to the one marked with the letter a are found contain- 
ing even a larger number of nuclei. These may be separated, as in 
the present instance, or they may be joined together by filamentary 
prolongations. The protoplasms around them contain haemoglobin. 
These cells undoubtedly represent ineffectual attempts at the forma- 
tion of red blood-cells. As will be seen, when we come to study the 
spleen, the successful formation of erythrocytes seems to take place 6 
GU1TE RAS, 
only from free nuclei similar to those found within this mass of proto- 
plasm. It will be seen, furthermore, that the successful formation of 
red blood-cells takes place only in contact with endothelial cells, and, 
in the present case, only in the spleen. We should not fail to notice 
the marked difference that there is between the cell at present de- 
scribed and the one containing pale nuclei. The nuclear filaments 
that may be found joining the fragments of the nucleus in the latter 
cells are pale and usually thick, whilst the filaments of the multi- 
nuclear erythroblasts are stained intensely red and are very fine. I 
should mention here that when the triple stain of safranin, orange G, 
and methyl-blue is used, the pale nucleus of the so-called leucoblasts 
takes a greenish color, whereas the homogeneous nucleus of the ery- 
throblasts, whether found free or within the protoplasm of large cells, 
takes the red stain of the safranin. In fact, the orange G acts as a 
decolorizing agent, removing the safranin stain from all the structures 
except the erythroblastic nuclei. In the upper part of Fig. 3 a few 
cells are shown presenting the greenish stain in contrast with a few 
homogeneous small nuclei stained red. The appearance of structure 
in some of these red nuclei will be noted later on. 
We find also in the bone-marrow a large number of cells contain- 
ing a large vacuole and presenting a few granules of fat in the proto- 
plasm around the vacuole. These are evidently fat-cells from which 
the infiltrated fat has been removed. 
Red blood-cells of various shapes are found in the channels of 
the marrow together with the cells above described. Many of them 
are very pale, and some of them are nucleated red blood-cells. We 
find nowhere any clear evidence of pigmentary degeneration of these 
cells. 
In the spleen we find a general increase of the connective-tissue 
reticulum. We find here, as in the marrow, a number of parallel 
channels, some of them lined with endothelium and others not. The 
latter are frequently crossed by fibrillar prolongations of the reticu- 
lum. The nodes of the latter present here and there stellate connec- 
tive-tissue cells and occasionally a flat endothelial cell with slightly 
projecting pale oval nucleus. These spaces contain cells similar to 
those that have been described in the bone-marrow. The number of 
red blood-cells, however, is greater, as is also that of the free nuclei 
that we shall describe here as true erythroblasts. These are most MARROW OF BONE AND SPLEEN IN LEUKEMIA. 
7 
numerous in the immediate vicinity of the channels that possess a 
distinct endothelial wall. These will be considered as capillaries. 
The point that I wish particularly to call attention to is the behavior 
of the erythroblasts in the immediate neighborhood of the endothelial 
channels. It is here that we see for the first time anything like the 
formation of red blood-globules. The process is something as follows: 
'file small free erythroblast sends a prolongation of varying length 
through the endothelial wall from without toward the interior of the 
vessel. This prolongation just within the endothelial lining swells 
into a bud which may be smaller or larger than the erythroblast from 
which it springs. This small bud presents the same peculiarities as 
the erythroblast outside. It is around this bud that we see the forma- 
tion of the red blood-cells. These seem to grow around the bud as a 
protoplasmic formation around the nucleus, except at the point of con- 
tact with the filament. If the bud presents a round shape, the haemo- 
globin-holding protoplasm is also globular. If the bud is pear-shaped, 
the protoplasm generally presents the same outline. Occasionally the 
protoplasm grows more from one side of the nucleus. The union be- 
tween the erythroblast and the bud may be very close, presenting the 
appearance of a large diplococcus divided by a line of endothelial 
membrane. In other cases the filamentary union may be as long 
as 9//. In some cases the bud has evidently been broken oft’, or has 
not formed, and the end of the filament can be seen pointing in the 
endothelial wall. When the bud-cells, with their haemoglobin envelope, 
are disengaged from the endothelial wall they constitute a nucleated 
red blood-cell. As some of these nuclear buds are found of very small 
size, with a large blood-cell around them, it is very probable that the 
latter may be loosened from its anchorage without any nuclear con- 
tents, thus constituting a fully developed red blood-cell. There is also 
evidence that the bud may be loosened from its moorings without any 
haemoglobin formation about it. In this case it constitutes, it appears 
to me, a true blood-plaque. The measurements of one of these buds 
gives for the red homogeneous nucleus within the endothelial channels 
3/i in diameter, the red body outside of the endothelial channel meas- 
ures 2fi in diameter, the filament between them and the haemoglobin 
capsule 5J/i. Another one in the same capillary presents a bud of 
H/* in diameter and a hsemoglobin capsule of 6//. The filament pene- 
trates the endothelial wall for a distance of 2//. Here it has evidently 8 
GUIT^RAS, 
broken off in the preparation. The outline of the haemoglobin envelope 
is generally globular, and the protoplasm is slightly granular and yellow. 
In another capillary we find the following evidences of budding. First, 
a homogeneous red body in diameter in close contact with the ex- 
ternal surface of the endothelial wall, through which it sends a tip-like 
projection 1 \fi long by 1 [i wide. This is capped within the blood- 
vessel by a pear-shaped red blood-cell very slightly larger than the 
erythroblast without. Next to this there is an endothelial nucleus, and 
just beyond it another filamentary figure (Fig. 5, a) with the following 
dimensions: Outside of the endothelial channel there is a round, in- 
tensely red homogeneous body. This body measures nearly S/i in diam- 
eter. There is no protoplasm about it. From this body a fine filament, 
slightly bent upon itself, extends a distance of 6/j. to a similar red body 
just within the endothelial line. The fine filament expands slightly 
as it goes through the endothelial wall, thus giving something of a 
pear-shape to the body within the endothelial channel. The latter 
body is equal in diameter to the one outside. The hmmoglobin cap- 
sule springs from one side of this body and is somewhat oval in shape, 
measuring 3 x 3\fi. A small portion of haemoglobin protoplasm is 
also 3een projecting from the opposite side of the endothelial bud. 
Further on we find (Fig. 5, b) a pear-shaped red body close to the endo- 
thelial lining, with a point imbedded in it. This is continued into a short 
filament somewhat bent upon itself, and terminates in a similar pear- 
shaped body within the endothelial channel. Springing from this we 
have a pear-shaped red blood-cell measuring about 6(« in diameter; 
next to this there is a closely joined couple of pear-shaped bodies. All 
these are found upon the same side of the capillary within a distance 
of 50fi. This process of filamentary budding never takes place through 
the endothelial nucleus, but through the cement or the protoplasm of 
the flat endothelial plate. 
There are several reasons to support the view that these filamentary 
buds project from without to the interior of the capillary. They are 
as follows : 
1. We always find an absence of haemoglobin around the erythro- 
blast outside of the endothelial channel. It is reasonable, then, to 
suppose that the globule acquires its hmmoglobin envelope after it 
penetrates the endothelium and comes in contact with the circulating 
plasma. 9 
MARROW OF BONE AND SPLEEN IN LEUKEMIA. 
2. We find in the majority of instances that the body outside of 
the endothelial channel is larger than the one within. 
3. We find also evidences of progressive chromatolysis in the bud, 
usually in proportion with the growth of the haemoglobin capsule. 
Sometimes there is nothing left but the point of the chromatin fila- 
ment to which the red blood-cell is attached. The bud further shows 
evidences of dissolution by breaking up into minute filaments radiating 
for a short distance from the point of attachment. The latter change 
gives these corpuscles, with their filamentary tail, the appearance at 
times of spermatozoa. 
The small, round erythroblasts are not always homogeneous. Some 
of them present within the well-stained nuclear matrix minute granules 
of a darker color. These are so arranged sometimes as to give the im- 
pression of mitotic figures. It is probable that they are such, but it is 
difficult to analyze them on account of the intense coloration of the matrix 
and the close packing of the supposed mi tons. In other cases we find 
within the erythroblast fine lines of a lighter color than the surround- 
ing substance. In such cases the chromatic substance appears broken 
up into four fragments by a pale crucial figure. The erythroblasts 
engaged in the process of filamentary budding are homogeneous, ex- 
cept that occasionally, as previously mentioned, the bud breaks up 
into small rays extending a short distance into the red blood-cell. 
Erythroblasts that are not engaged in the process of filamentary bud- 
ding are frequently of a larger size. In these a densely packed coil 
of chromatin may be seen, and sometimes a rearrangement of the 
miton in obscure karyokinetic figures. I am of opinion that in such 
cells the nuclear matrix may swell before the nucleus divides, and the 
cell is thus converted into the so-called leucoblast. 
Though the splenic channels, endothelial or non-endothelial, are 
frequently crowded with red blood-cells of varying shape, we find very 
little evidence of pigmentary degeneration. The irregular distribu- 
tion of these red blood-cells makes it difficult to decide upon the nature 
of the endothelial channels—that is, as to whether they are or are not 
true capillary continuations of the bloodvessels of the spleen. In my 
opinion they are; but this question belongs to one of the many diffi- 
cult problems in the histology of the spleen. 
For purposes of comparison, the spleens of a case of cardiac amemia 
and of pernicious progressive anaemia have been examined. In neither 10 
GUITE HAS, 
of these have the filamentary figures been found. It must be stated, 
however, that in the case of pernicious ansemia the tissues were not 
fixed in osmic acid, and that in the case of cardiac ansemia, though the 
attempt Avas made to fix the tissues by means of osmic acid, the result 
was not quite successful, evidently on account of post-mortem changes. 
In the spleen of cardiac anmmia the number of red blood-cells in 
the spleen channels is very great. The endothelial cells show no evi- 
dences of division. They are, however, more numerous and more 
globular than in the cases of leukmmia. They project further into 
the endothelial channel. Sometimes they are very close together, 
leaving scarcely any space between them. The closer together they 
lie the more prominent and globular they appear. These endothelial 
nuclei measure between 9// and 12// in length and 5// in width. The 
endothelial channels, besides containing frequently a large amount of 
hmmoglobin-cells, contain also a few leucocytes. The majority of 
them are multinuclear, and many of them contain fat-granules in 
their protoplasm. Occasionally we find an erythroblast nucleus in the 
interior of a protoplasmic body containing haemoglobin. These are 
not found in endothelial channels. Free erythroblasts are also nume- 
rous. 
In the case of pernicious anaemia the endothelial nuclei are still 
more swollen and numerous than in the case of cardiac anaemia. The 
capillary spaces are evidently dilated. In both the spleen and mar- 
row we find distinct evidences of pigmentary degeneration of the red 
blood-cells. We also find, especially in the marrow, rows of fat-gran- 
ules in the endothelial lining of the vessels. These rows are generally 
discrete, and sometimes they bifurcate, leaving the nucleus in the 
middle. The number of erythroblasts is smaller in these organs than 
in those of leukaemia. The same may be said of the number of eosino- 
phile cells. 
As to the origin of the erythroblasts, the opinion is maintained in 
the course of the paper that all the cells described may develop one 
from the other through a species of metaplasia, the starting-point of 
which is probably the lymphocyte. The erythroblast appears to 
originate in the shape of spherical fragments of chromatin that are 
thrown off from the nuclei of other cells; and the latter may again 
develop from the erythroblasts by a process of chromatolysis and 
swelling of the nuclear matrix. I do not pretend, however, to insist MARROW OF BONE AND SPLEEN IN LEUKEMIA. 
11 
upon this point. The opposite view, that gives the erythroblasts a 
specific character and traces their origin to embryonal red blood-cells, 
is also well supported. 
My object in this paper has been to call attention to the process of 
filamentary budding of erythroblasts through the endothelial wall as 
a factor in the formation of red blood-cells. It is very improbable 
that this process should occur in the leukmmic spleen and not under 
normal circumstances.1 
Figs. 4, 5, 0, 7 show capillaries of the spleen in which the pro- 
cess of filamentary budding manifests itself. Fig. 5 has been de- 
scribed in the text. In Fig. 6 a couple is shown in which the endothe- 
lial bud is of very small size and shows evidences of breaking up into 
indistinct filaments within the red blood-cell. This figure shows an 
arrangement of the parts that is frequently met with, namely, that 
the endothelial channels are much narrower than the non-endothelial 
channels around them. Fig. 7 shows a very long filamentary bud. 
In Fig. 4 a capillary is shown, in the upper border of which three 
filamentary buds are forming. 
Figs. 8 and 9 represent peculiar bodies found in the marrow of the 
case of leukaemia. It is impossible to account for these either by sup- 
posing them to be forms of degeneration of the nucleus or groups of 
red blood-cells. They resemble somewhat a sporoblast formation. By 
careful focussing the group is found to consist of nine cyst-like bodies, 
each one of which measures 3-4/i. They are vesicular in structure 
and present a rod-shaped body lying near the periphery of the vesicle. 
The vesicles take a pale reddish-gray color with safranin, and the rod- 
shaped body stains intensely red. A considerable number of these 
are found in the marrow of the case of leukaemia. They may be found 
in groups of twos and threes; the one represented in Fig. 9 and repro- 
duced from a photograph in Fig. 8, is the largest and most distinct of 
these groups that I have found. It is possible that these may be indi- 
cations of the so-called “vesicular degenerations” of the nucleus, or 
that they might be red blood-cells of small size presenting some change 
in their wall which leads to the appearance of the rod-shaped nucleus. 
It must be admitted, however, that these explanations are not satis- 
factory and that the nature of these bodies must be left in doubt until 
opportunity is given for a more complete investigation. 
1 Since the reading of this paper I have found the same process in the spleen of the rat, 
and of the rabbit, after hemorrhage. 12 
MARROW OF BONE AND SPLEEN IN LEUKEMIA. 
EXPLANATION OF ILLUSTRATIONS. 
Fig. 1. Bone-marrow of leukaemia, large irregular nuclei, with haemoglobin 
protoplasm. Compen. oc., No. 4. Apo. oil imm., 2 mm. 
Fig. 2. Bone-marrow of leukaemia, a. Large polymorphous nucleus with 
haemoglobin protoplasm, b. Large erythroblasts with imperfect mitotic division 
of the nucleus, c. Fat-cells. Compen. oc., No. 8. Apo. oil imm., 2 mm. 
Fig. 3. Bone-marrow of leukaemia, a. True erythroblasts surrounded by a 
mass of protoplasm containing haemoglobin. Compen. oc., No. 4. Apo. oil imm. 
2 mm. 
Fig. 4. Capillary of leukaemic spleen, showing the budding of erythroblasts 
through the endothelial wall. Three specimens are shown in the upper wall. 
Compen. oc., No. 4. Apo. oil imm. 2 mm. 
Fig. 5. Capillary of spleen, described in detail in the paper, a and b indi- 
cate erythroblasts in process of filamentary budding. Case of leukaemia. Pro- 
ject. oc., No. 2. Apo. oil imm., 2 mm.; camera length, 85 cm. 
Fig. 6. Section of leukaemic spleen, showing erythroblasts in process of bud- 
ding. Minute bud within the endothelial channel surrounded by a haemoglobin 
envelope. Project oc., No. 2. Apo. oil imm., 2 mm.; camera length, 85 cm. 
Fig. 7. Long filamentary budding in endothelial channel. Project oc., No. 2. 
Apo. oil imm., 2 mm.; camera length, 85 cm. 
Fig. 8. a. Sporoblast-like body, bone-marrow of leukaemia. Project oc., 
No. 2. Apo. imm., 2 mm.; camera length, 85 cm. 
Fig. 9. Sporoblast-like body. Compen. oc., No. 4. Apo. oil imm., 2 mm. Fig. 1. 
Fig. 2. 
Fig. 8. Fig. 4. 
Fio. 9. Fig. 5. 
Fig. 6. 
Fig. 7. 
Fig. 8.