ON THE SOCALLED “NEW ELEMENT” OF THE 
BLOOD AND ITS RELATION TO COAGULA- 
TION. By GEO. T. KEMP, A. B., Fellow in Biology, Johns 
Hopkins University. With Plate XIX. 
Whenever a figure appears in heavy type, the number refers to the corre- 
sponding number in the Bibliography under the name of the author. 
In 1878 Hayem called attention to an element of the blood, 
whose existence and significance seem to have been almost 
wholly unrecognized. Hayem called these elements hcematoblasts, 
and endeavored to show that they were early stages in the devel- 
opment of the red corpuscles. He also pointed out that they 
probably are connected in some way with the coagulation of 
blood ; but as the theory of their hsematoblastic function did 
not meet with general acceptance, the whole work seems to have 
been passed over with less attention than it merited. 
Much greater prominence was given to the subject, when, in 
1881, Bizzozero claimed independent discovery of the same 
elements, and emphasized particularly their connection with 
coagulation. He attacked the theory then in vogue as to the 
relation of the leucocytes to coagulation, and called forth many 
severe criticisms, which in turn were answered ; thus the ques- 
tion was brought prominently to the attention of histologists. 
The result was several important investigations ; but the irrecon- 
cilable results obtained by different observers, rendered it 
desirable that the whole matter should be made the object of 
further research. I was therefore led to take up the study of 
this question, and the results of my work thus far, form the 
material for this paper. 
Before going into a discussion of the subject, which will 
involve the results of other observers, I will give a brief descrip- 
tion of the “ new element ” as observed by myself, in order that 
what is to follow may be better understood. 
In describing it I shall adopt the French name jplaque, as 
it is short, more or less appropriate, and not liable to be confused 294 
GEO. T. KEMP. 
with any other name which may be used in connection with this 
subject. 
If a drop of osmic acid be placed on the finger, and the finger 
pricked with a needle through the drop, the elements of the 
blood will all be hardened and preserved in their natural appear- 
ance immediately upon leaving the vessel. 
If now a thin film of the blood mixed with osmic acid be ex- 
amined under the microscope with a good lens magnifying about 
600 to 800 diameters, the plaques may be seen floating in the 
plasma among the red corpuscles and leucocytes. 
They are very pale and homogeneous structures, varying 
greatly in size, but mostly about one-third or one-fourth the 
diameter of the red corpuscles. When seen in surface view, they 
may appear either circular or elliptical, and seem at first sight to 
be flat, but a more careful observation with a fine objective, will 
reveal the fact that they are biconcave, although not as much so 
as the red corpuscles. (See Plate, Fig. 1, and photograph.) This 
is more plainly marked when we examine them seen on edge, 
in which case they show the characteristic dumb-bell shape pre- 
sented by biconcave bodies when viewed in this position. 
The form of the plaques, when hardened by the above method, 
never undergoes change. This is not the case, however, in blood 
drawn and allowed to clot. To study the plaques under these 
circumstances, the following method may be adopted: The 
finger is pricked and a good-sized drop of blood squeezed 
out. This is taken immediately upon a cover-slip, and then as 
quickly as possible most of it is washed otf by a jet of .75 per 
cent. NaCl solution from a wash-bottle. The cover-slip is now 
placed on a slide, and transferred to the microscope stage with as 
little loss of time as possible. The plaques have the property of 
sticking to the slip, while the other elements are easily washed 
away by the jet, so that upon examination the whole field will 
be seen to be filled with plaques, some of them isolated, but 
most of them grouped in masses consisting of from two or three 
to a dozen or more. (Plate, Figs. 2 and 3, and photograph.) 
They are now not pale and homogeneous, with a symmetrical 
outline, but appear glistening and granular, and their contour, 
instead of being regularly oval or circular, has become jagged. 
These changes are the more marked the longer the time which has “NEW ELEMENT” OF THE BLOOD. 
295 
elapsed before the preparation is observed; and they may be seen 
to take place step by step while a preparation is being watched. 
The form of the plaques continues to undergo change, until 
finally, where they are grouped together, only a granular mass 
is found, in which the individual plaques can no longer be clearly 
distinguished. (Plate, Figs. 4 and 5.) Pari passu with these 
changes, processes are seen which run out from the granular 
masses ; and when coagulation sets in, these processes are nearly 
always found to be continuous with threads of fibrin. (Plate, 
Figs. 5 and 6.) 
The threads of fibrin are sometimes deposited as long needle- 
shaped crystalloids which are often seen lying free in the field, 
and not in connection with the granular masses; but the greater 
number are formed most thickly around these masses, from which 
they often radiate as centres. (Plate, Figs. 5, 6 and 7.) 
If too much of the blood has been washed away by the jet of 
salt solution, no formation of fibrin will take place; if there is 
any part of the field where the blood still remains thick, the best 
observations can be made on the edges of this area. 
The preceding description, I think, is given in sufficient detail 
to indicate what I mean by the term plaques. 
§1. 
Among the earlier observers, Muller, Mandl, Henle, Wharton 
Jones and others, have described various colorless elements 
existing in the blood; and it is not unlikely that they saw the 
plaques, more or less modified ; but their descriptions are not 
sufficiently minute to enable us to make a positive decision. 
Gerber1 has described in the blood of mammals, bodies which 
he calls free nuclei, having a diameter of jfo line (— 4.1 /u). 
Arnold2 and Andral have also been recognized as possible dis- 
coverers of the plaques. 
Historical. 
1 Gerber. Allgemeine Anatomie. Quoted by Zimmerman, 2, 226. 
2 Arnold (Anatomie, Band I, p. 181) has been referred to by Zimmerman (2, 
226) and Hayem (8, 581) as having probably seen the plaques. Hayem gives 
Arnold’s measurement as foot (which I take to be a misprint) and Zim- 
merman gives it as zoll (= about 16.7 /*), which is entirely too large for 
a plaque. Zimmerman thinks that the structures described by Andral (Essai 296 
GEO. T. KEMP. 
Simon,1 in studying blood drawn into a solution of potassium 
ferrocyanide, found small bodies which he took to be molecules 
of fibrin, but which were probably plaques, and other elements 
deformed by the reagents that he used. 
Donn62 (1842) described and figured the plaques so as to leave 
no doubt as to their identity. He calls them globulins, and says 
they belong properly to the chyle, from which they are supplied 
to the blood. He describes them as small white particles, or 
little rounded grains, isolated or irregularly agglomerated, with 
a diameter of not more than millimeter (=z 3.3 //), and 
says they are important as being the “ premiers 6l6ments des 
globules sanguins.” He also noticed the tendency of the 
plaques to adhere to each other. 
The globulins of Donn6 are not the same as the globulins of 
Milne-Edwards, nor of Robin; the former using the word to 
denote small, fatty particles found in the blood especially after a 
meal, and the latter using the word with reference to the small- 
est variety of leucocytes.3 
To Zimmerman (1846) belongs the credit of first having studied 
the plaques with the aid of micro-chemical reagents. He (2, 227) 
repeated the experiments of Simon by drawing blood into a solu- 
tion of potassium ferrocyanide, which prevents clotting, and then 
examining with the microscope. He found small bodies, occur- 
ring “ by the billion,” which he describes as quite colorless and 
more or less strongly refractive; they have not a well-defined 
contour, and 'when out of focus appear as dark points. They 
vary in size from t° Tihr °f a ne (— about 2 fi to 5.2 /1). 
He does not agree with Simon that they are molecules of fibrin, 
and calls them “ Elementarkorperchen,” because they cannot be 
d’hematologie pathologique, p. 32), as molecules of fibrin are the same as those 
mentioned by Arnold, while Hayem (8, 582) thinks that Andral’s molecules of 
fibrin are likely plaques, because of their small size. I was not able to refer to 
the papers of Arnold and Andral, so that I cannot reconcile the statements of 
Zimmerman and Hayem. Possibly the measurements referred to granular 
masses, which Arnold may have explained, without the explanation being quoted 
by Zimmerman or Hayem. 
1 Simon. Anthropochemie. 
2 Donne. Comptes Rendus de l’Academie des Sciences, 1842 (Vol. 14, pp. 
366-368), and Cours de Microscopie, etc., 1844, p. 85 ; and Atlas, Plate 6, 
Pig. 21. 
3 See Hayem, 8, 583. “NEW ELEMENT” OF THE BLOOD. 
297 
broken down into any other smaller-formed element, and because, 
as he claims, they have power to grow by intussusception and 
form new cells. As to their origin, he supposes that they, like 
the white corpuscles, belong properly to the lymph, from which 
they are supplied to the blood. 
It is more than probable that under the name of elementary 
corpuscles, Zimmerman has included not only the plaques, but 
also other bodies, especially certain granules often found in the 
blood, which are partly normal and partly the result of some 
methods of preparation. He speaks of elementary corpuscles as 
occurring in the blood of a frog; and describes in addition the 
nucleated plaque (found by later observers in the blood of all 
ovipara) as an intermediate stage in the formation of red cor- 
puscles from the elementary corpuscles. He also claims to have 
seen the elementary corpuscles in defibrinated blood. 
These inaccuracies have led some histologists to question 
whether Zimmerman really found anything but detritus, resulting 
from his methods of preparation; but his description of the 
behavior of the elementary corpuscles with certain reagents— 
e. g., water, caustic potash, salt solutions, etc.—corresponds so 
nearly with the reactions of the plaques under similar conditions, 
that we can hardly doubt that he had plaques under observation. 
Beale (1864), (1, 42) describes as existing in the blood, 
“numerous corpuscles differing very much from the ordinary 
red corpuscles in size, color, and refractive power. They are 
much smaller than the latter; they exhibit a granular appear- 
ance and are colorless. They might be described as small white 
corpuscles, but many are much smoother than the colorless 
corpuscles. It is not easy to see these corpuscles unless the 
blood is examined by powers magnifying upwards of 1000 
diameters. Such corpuscles are exceedingly faint, and can only 
be distinguished if great care be employed The small, 
faintly granular corpuscles are colored by carmine,x while the 
ordinary red corpuscles .... are not.” 
At first, from the above description, it seems as if Beale had 
reference to the plaques; but the fact that the structures he 
describes stain with carmine excludes this supposition. 
Beale (2,48) also describes other structures which occur “only 
1 The italics are mine. 298 
GEO. T. KEMP. 
in the blood of man and the higher animals,” and which are of 
the same refractive power as the white corpuscles. These 
particles he thinks may develop into white or red corpuscles. 
He further mentions (2, 60) small bodies, which he believes to 
be a kind of white corpuscle, which stick to the slide, while the 
red corpuscles are drawn hither and thither by the current. 
These descriptions may well apply to the plaques. 
The “granular masses” described by Max Schultze (1865), 
which have been more or less well known to histologists ever 
since, are undoubtedly plaques, as may be seen from his descrip- 
tion. He says (1, 36 and 37), “I find more or less plentiful 
in my blood and the blood of many others, . . . clumps of color- 
less balls varying in size accordingly as they are made up of few 
or many individuals. The latter measure .001—.002 min., and 
occur also singly in the blood, but more frequently they are 
found sticking together to form an irregular, finely granular mass. 
The balls are quite colorless, homogeneous or finely granular and 
pale, .... and hence, as also on account of their small size, they 
can only be made out with a good lens. 
“They are, however, not always regularly ball-shaped: they 
are often angular and drawn out, and then they present a sharper 
contour and a clearly granular appearance It looks as 
though they may have come from the leucocytes, but of this we 
must remain uncertain as long as we are in the dark as to the 
fate of the latter.” 
He also obtained the characteristic reactions of plaques with 
water, acetic acid and other reagents. 
Eiess (1872), (1, 240) examined blood drawn by venesection, 
with as little loss of time as possible, and found “ Stabchen und 
belle Kiigelchen von ahnlichem Glanze mit den farblosen Kor- 
perchen.” 
The “Kiigelchen” are mostly round, but sometimes angular. 
They are usually about y the diameter of red corpuscles (.7-1.5 /1), 
but they vary from very minute bodies to half the size of the red 
corpuscle. This last statement makes it somewhat doubtful 
whether Riess has made the necessary distinction between the 
plaques and the smaller varieties of granules found in the blood. 
He also claims to have found all stages between leucocytes and 
granular masses of plaques. “NEW ELEMENT” OF THE BLOOD. 
299 
In 1873 Ranvier laid a communication (2 and 31) before the 
Soci6te de Biologie, on the “Formation of Fibrin in Blood 
Removed from the Vessels.” He describes the threads of fibrin as 
radiating from granules, or groups of granules, which he thinks 
are chemically identical with fibrin. From references made to 
these granules in later publications we know that he had refer- 
ence to the plaques, which he has also seen free (1, 215). 
In connection with Ranvier’s communication, Vulpian (1, 91) 
gave the results of some observations which he made on blood. 
He examined the blood of many persons, and found that there 
were always present small corpuscles, either singly or in groups, 
which quickly adhered to the slide, and were not carried about 
by currents, like the other elements. When coagulation takes 
place, the threads of fibrin are often seen to radiate from prolon- 
gation of these corpuscles, or from the edge of groups into which 
they have become collected. He identifies these corpuscles with 
those described by Riess, and takes occasion to combat the state- 
ment of Riess that they are the result of pathological conditions, 
maintaining that he has found them in the blood of healthy 
persons as well as in the blood of the sick. 
Bearing also on the relation of the plaques to the fibrinous 
network are the observations of Nedsvetzski (1, 147-150), 
made independently of the work of Ranvier and Vulpian, and 
published about the same time. He describes small, homoge- 
neous bodies about the size of the granules of the leucocytes, 
which he regards as normal constituents of the blood, and gives 
them the name of “ Blutkornchen ” or “ Hsemococci.” 
After fibrin has been formed, these small bodies are found like 
knots in the meshes of the network. He also says that these 
bodies possess the power of movement. 
The Proceedings of the Royal Society of London for 1874 
contain “ An Account of Certain Organisms Occurring in the 
Liquor Sanguinis,” by Osier, in which he describes certain inter- 
esting changes which take place in the granular masses when 
heated in serum to 37°C. In the same communication Osier 
secures for himself the credit of having first discovered the 
plaques in the bloodvessels. His observations were made on 
connective tissue from the back of young rats, in the vessels of 
which the plaques were seen scattered freely in the blood among 300 
GEO. T. KEMP. 
its other elements. The figures which accompany Osier’s article 
are very good. 
From this time until the publication of Hayem’s work, the 
only literature bearing on the subject are the papers by Schmidt 
(2 and 3), Semner (1) and Boettscher (1). Schmidt and Sem- 
ner examined “ the granular masses of Max Schultze,” and 
described them as being derived from a peculiar corpuscle not at 
all like the plaques. 
Boettscher (1, 298 and plate), who believes the red corpuscles 
to be nucleated, lias figured structures found by him among the 
red corpuscles, which he regards as nuclei set free from the latter ; 
but which, judging from his figures, are probably plaques. 
To show what a confusion of ideas prevailed in regard to this 
subject before Hayem’s work came out, I cannot do better than 
follow Hayem in quoting from Kobin {Lemons sur les humeurs 
normals et jpathologiques, 1874). After reviewing the work of 
Biess and Vulpian, he says: “II est de fait qu’avec un peu 
d’attention on trouve de ces globules sur presque tons les sujets, 
bien qu’en tres petit nombre. On en voit mOne circuler de loin 
en loin entre les hematies et les leucocytes, sur les batraciens et 
les poissons vivants. Ces corpuscules se trouventsoit isoles, soit 
reunis en petits groupes. Ils sont larges de 0.002 a 0.006, c’est- 
a-dire que certains atteignent presque le volume des hematies et 
meme des leucocytes. Ils sont incolores, hyalins, homogenes ou 
a peine grenus. Parmi ceux-ci il en est qui se deforme lente- 
ment comme les leucocytes. Pour ces derniers l’action de l’eau 
et de l’acid ae6tique sous un fort grossissement montre aisement 
qu’il ne s’agit la que des leucocytes encore petits et en voie 
devolution, qu’il s’agit en d’autres termes de ceux des leucocytes 
que M. Donn6 (1844) a sous le nom de globulins, larges 
de 0.003.” 
With Hayem’s work (1878) begins a new era in the history of 
the plaques, for, by the careful and thorough investigation to 
which he subjected them, he was not only able to remove all 
doubts as to their existence and identity, but he also traced their 
close connection with the process of coagulation, and has ad- 
vanced some very suggestive speculations regarding their affini- 
ties to the red corpuscles. 
Bizzozero (1881 and 1882) so confirmed and extended the “NEW ELEMENT” OF THE BLOOD. 
301 
observations of Hayem by new and ingenious methods, that all 
the work done later, has been carried on with a more definite 
understanding of the subject, and the results of the observations 
may be considered as confirming or conflicting with those of 
Hayem or Bizzozero. 
Before concluding the historical chapter of this article, the 
claim of Norris to the discovery of the plaques should be con- 
sidered. 
Norris bases his claim on a communication read before the 
Birmingham Philosophical Society, in 1879, in which he says 
(2, 163): “ There exist in the blood of mammalia .... color- 
less, transparent, biconcave discs of the same size1 as the red ones. 
Between these two kinds of biconcave discs others are demon- 
strable having every intermediate grade of color} . . . They 
often present themselves as small spheres, at other times as discs, 
and at others, liquid-like, they take the shape of the interstices- 
in which they lie.” 
In a later article (3, 562), Norris speaks of Hayem’s haemato- 
blasts as fragments of one form of his corpuscles. 
From a comparison of these descriptions with those given at 
the beginning of this paper, I think it will appear evident that 
what Norris saw was not the plaques, but red corpuscles which 
had lost their haemoglobin. 
§ 2. 
Histology of the Plagues in the Blood of Mammals, with Methods- 
of Preparation. 
The plaques have been found under different circumstances by. 
so many observers (see historical section), that it is now needless 
to enter into a discussion as to their presence in the blood—at 
least after it has been removed from the bloodvessels. 
The method of preparation may be varied to advantage accord- 
ing to the point which we wish especially to make out. The 
following may be recommended as giving very satisfactory 
results:— 
A drop of the preservative fluid is placed on the finger, and 
the finger pricked through the drop, so that the blood may come 
1 The italics are mine. 302 
GEO. T. KEMP. 
in contact with the reagent immediately upon leaving the vessel. 
The drop of blood thus mixed with the reagent is next taken on 
a cover-slip, and the latter laid on a slide and examined. To 
prevent evaporation, melted paraffin may be painted around the 
edges of the slip. 
The part of the finger from which the blood is to be drawn 
should be washed with water, alcohol and ether, in the order 
name„d. 
"Whenever it is necessary to run any fluid under the slip by 
suction writh filter-paper, it can most easily be accomplished by 
putting the minutest part of a drop of very thick balsam on each 
of the four corners of the cover-slip. When laid on the slide the 
balsam should be on the under side of slip. By this arrangement 
the slip is held to the slide by the balsam, and will not be floated 
about by an excess of the fluid run under. It also prevents the 
slip from lying so close to the slide as to interfere with the filter- 
paper readily absorbing the fluid. Other observers have used 
paraffin, but I find thick balsam rather superior. 
On collecting blood directly from a needle-prick into a suitable 
liquid, the plaques are preserved in their original form, but their 
relative number is small, and they do not readily attract atten- 
tion lying among the more prominent corpuscles. Their pres- 
ence in the blood may be demonstrated more strikingly by pro- 
ceeding as follows : A large drop of blood is squeezed from the 
finger and taken directly on a cover-slip. The surface of the 
cover-slip to which the blood is adhering, should next be touched 
three or four times to the surface of some .75 per cent, salt solu- 
tion in a watch-glass, or washed with a jet of the same solution 
from a wash-bottle, until nearly all of the blood is apparently 
washed off. This part of the proceeding should be finished as 
quickly as possible, and the cover-slip placed in a watch-glass of 
osmic acid for about twenty to thirty minutes or longer. It 
should then be washed by leaving it in a watch-glass of water 
for several minutes, after which it may be mounted in acetate of 
potash and examined. 
The plaques are seen in great numbers all over the field. 
Sometimes they may be found singly, but more frequently they 
are seen in groups, the size of which depends upon the length of 
time between drawing the blood, and getting it into the osmic “NEW ELEMENT’ OF THE BLOOD. 
303 
acid. (See photograph.) If this he very short, the masses are 
small, and the individual plaques of which they are composed 
may be plainly distinguished; but the longer the time, the larger, 
as a rule, will be the masses, and the less distinctly can the 
individual plaques be seen. 
For hardening and preserving the plaques, I think 1 per cent, 
osmic acid is the best reagent that can be employed. Instead of 
osmic acid we may use Hayem’s solution, the formula for which 
is:— Distilled water, 200 
Nad 1 
NasSCh 5 
HgCL 0.5 
I prefer, when using Hayem’s solution, to dilute it with its 
volume of 75 per cent. NaCI solution. This avoids the disad- 
vantage of a precipitate if the fluid should evaporate or become 
more concentrated, and preserves the plaques very well. The 
plaques show very plainly when examined in Hayem’s solution, 
their appearance being more striking than when seen in osmic 
acid; but we are more apt to get very slight irregularities of 
outline, together with minute granules deposited on the face of 
the plaque, which are less frequently seen in osmic acid prepara- 
tions. This is probably due to Hayem’s fluid acting somewhat 
less quickly than osmic acid, thus allowing to appear the first 
changes which the plaques undergo in breaking down. I have 
also tried Bizzozero’s fluid, and find it an admirable medium for 
studying the changes undergone by the plaques in breaking 
down, as the changes go on very slowly in this solution. Bizzo- 
zero’s fluid is .75 per cent. NaCl solution, to which methyl violet 
is added. He recommends the ratio of methyl violet to salt 
solution 1: 5000. I did not determine the ratio of the violet to 
the salt solution in the fluid which I found to give the best 
results, but I think it contained relatively more of the staining 
element than that recommended by Bizzozero. 
I have also trifid the method of preservation by drying. 
Preparations were dried both by spontaneous evaporation and 
carefully over an alcohol flame, but the results of this method 
were so far inferior to those obtained by hardening in osmic acid 
or Hayem’s fluid, that I did not employ it to any great extent. 
If the blood is first hardened in osmic acid and then dried, the 304 
GEO. T. KEMP. 
results are much more satisfactory, although .then the drying 
process is unnecessary. 
As mounting media, I have tried the following:—Canada 
balsam, Dammar varnish, glycerine, glucose, acetate of potash 
(saturated solution) and Hayem’s fluid. 
Balsam and Dammar varnish were generally used with the 
dry specimens. They may either be mounted directly (if per- 
fectly dry), or passed through turpentine or xylol. The latter 
method is rather preferable. Balsam and Dammar were also 
often used with specimens stained in an alcoholic fluid, but in 
general it is best to avoid the use of alcohol, as it appears to 
cause a slight shrinkage even after hardening with osmic acid. 
Glycerine is not the best medium for mounting unstained 
plaques, especially when hardened in osmic acid, as they lose 
somewhat in clearness. For stained specimens it works very well. 
Glucose makes a very satisfactory temporary mounting 
material. It is used in concentrated aqueous solution, and has 
the advantage of becoming perfectly hard and requiring no 
cement.1 
Acetate of potash is probably the best mounting medium of 
all those employed. In it both plaques and fibrin threads stand 
out clearly and sharply defined. 
In addition to these regular mounting media, the plaques may 
be preserved for a time in Hayem’s fluid. This does very well 
for some weeks, but in specimens several months old there is 
often a coarsely granular precipitate, apparently from the plasma, 
and sometimes crystals are deposited. 
The description which I have given of the plaques in the 
introductory pages of this article, agrees for the most part with 
the descriptions of previous observers, but there are some points 
of difference to which it is desirable to call attention. 
In the first place, their biconcavity is still a subject of dispute. 
Bizzozero (7, 18 and 3, 357) asserts most positively that they are 
flat, but says they may become biconcave when treated with 
strong salt solution; while Lowit (3, 108) says that they are 
1 This was found to hold good for several months, but in most of the speci- 
mens examined after the lapse of a very hot summer, the glucose was found to 
have crystallized out, and the specimen consequently was ruined. This method 
cannot, therefore, be recommended for permanent preparations. “NEW ELEMENT” OF THE BLOOD. 
305 
never biconcave in blood drawn into salt solutions, but that 
under certain conditions a biconcavity may be seen in them 
when examined in peptone blood. Hayem (5, 695 and 697) 
and Laker (1, 177), on the other hand, uphold the view that they 
are normally biconcave, while Schitnmelbusch (2, 217) says they 
are not biconcave when examined circulating in the capillaries 
of the mesentery, but become so when the blood is drawn into 
Hayem’s fluid. As I have never examined the circulation in 
the mesentery, I am not prepared to endorse or dispute this 
statement; but I can say that in addition to seeing them on edge 
and making out their characteristic dumb-bell shape, I have suc- 
ceeded several times in observing them rolling over and over in 
a very slow current, so that at least in osmic acid and Hayem’s 
solution I was able to convince myself beyond all question of 
their biconcavity. The biconcavity also appears in the photo- 
graph, which was made from a specimen stained -with Bismarck 
brown. 
Bizzozero also describes the plaques as granular, and not homo- 
geneous. This is due, I think, to the method which he employed. 
If he had drawn the blood directly into osmic acid, he would 
have found the plaques homogeneous. 
The sizes, as given by different observers, vary within wide 
limits, but the mean of each of them generally corresponds pretty 
well with the mean of the others. 
Osier gives their measurement as one-eighth to one-half of 
that of the red corpuscles, which I think expresses with sufficient 
accuracy their extreme variations in size. Hayem (5, 703) at 
first describes the plaques as homogeneous and non-nucleated, 
but later (17, 480 and 481) he changes his opinion and states 
that they contain a nucleus, which may be brought into view by 
staining with haematoxylin. He gets his best results from dry- 
ing a film of blood on the slip—a method which I have never 
considered as free from objection, especially when heat is applied 
to effect the desiccation. 
In a later article (18, 372) he again reiterates his statement 
in regard to the nucleus of the plaques, and this time with more 
confidence than before. At the same time, he expresses himself 
in favor of the view that red corpuscles are also nucleated. 
Afanasef investigated this question after Hayem, and comes 306 
GEO. T. KEMP. 
to the conclusion that what Hayem regards as a nucleus is really 
only a precipitation of granules in the centre of the plaque. 
Schimmelbusch (2, 225) takes a view very similar to that of 
Afanasef. I have some very satisfactory specimens of plaques 
stained with haematoxylin ; some are mounted in acetate of 
potash, others in balsam, but none were preserved by the drying 
method. A thorough examination of these preparations has 
failed to reveal a nucleus, but the characteristic appearance of a 
concave body when seen full on the surface was presented. This 
was rendered more prominent than in the unstained plaque, and 
in so small a structure, it is possible that an appearance of this 
sort may have been taken for a nucleus. Preparations stained 
with Bismarck brown and magenta show the same thing. 
The same question as to a cell-membrane may be raised with 
regard to the plaques, as in the case of the red corpuscles. 
The plaques, under certain circumstances, will shrivel and 
become crenate, just as the red corpuscles; and it is quite a usual 
thing to see both plaques and red corpuscles caught at one end 
and drawn out by a current so as to present a long, pear-shaped 
or even threadlike appearance. (Plate XIX, Fig. 1, d.) 
Hayem (5, 696 and 717) described the plaques as containing 
haemoglobin. In this he was supported by Mayet (1, 213 and 
241), and by the work of Laptschinsky (1, 658), who describes 
corpuscles about one-third the size of the red corpuscles, which 
are sometimes more strongly colored than usual, and sometimes 
colorless. 
Hlava also (1, 403) has seen circulating in the mesentery of 
rabbits, small, mostly round elements colored by haemoglobin, 
which are either young red corpuscles (which he doubts) or 
fragments of old ones. These are not the true plaques, for he 
has described the latter 404), and says they are colorless. 
Eiess (1, 224), Bizzozero (3, 352), Laker (1, 200), Hlava (1, 
403), Halla (1, 220), and Schimmelbusch (2, 217), all take the 
view that the plaques are without color. 
Laker has noticed the “ teinte verdatre ” of Hayem, but 
thinks it only a shimmer from the surface. He has seen the 
same thing in the leucocytes. Another proof that the color is 
not proper to the plaques, is the fact that when seen in groups, 
one lying over the other, the color is no more intense than when “NEW ELEMENT” OF THE BLOOD. 
307 
seen singly or isolated. This has also been noticed by Laker 
(1, 179), and by Halla (1, 220), as well as by myself. 
Another interesting relation between the plaqnes and red 
corpuscles has been pointed out by Hayem (5, 728), Bizzozero 
(3, 357), and Lavdovsky (2, 64). They find that in different 
animals the size of the plaques always varies in the same ratio 
as that of the red corpuscles. 
To these many interesting points of relationship between the 
plaques and red corpuscles, I can add that, in trying to get a 
preparation from which to make a photograph of the plaques, 1 
tried ten different staining fluids on the blood after hardening in 
osmic acid, and in every case the plaques and red corpuscles 
stained proportionately. 
In addition to the points of resemblance already given, 
Hayem (10, 565) also mentions a somewhat complicated process 
of drying, -washing and staining in which the plaques deport 
themselves exactly like the red corpuscles. 
The many prominent marks of relationship between the 
plaques and red corpuscles, together with the idea that they con- 
tained haemoglobin, prompted Hayem to advance his haemato- 
blast theory : that the plaques are early stages in development of 
the red corpuscles, the gap between the largest plaques and 
smallest red corpuscles being filled by elements to which Hayem 
gives the name “ globules nains.” In support of his haemato- 
blast theory, Hayem (11,120) has found that in pathological 
conditions of the system where “ new blood ” is demanded, we 
always find the plaques in increased numbers— a fact which had 
previously been stated by Kiess (2, 696). 
The sources of error in all numerical determinations of the 
plaques are very great, on account of the tendency which these 
elements have to adhere to each other or to any foreign body 
with which they come in contact. I was early led to see that 
for this reason numerical determinations where the blood had to 
be measured, were practically of little value. The same point 
has been made by Schimmelbusch (2, 226-231), who, in addition 
to the results of his own observation, has shown that enormous 
variations exist in the results of the same observer, while differ- 
ent observers have obtained diametrically opposite results, in 
determining the number of plaques in certain pathological con- 308 
GEO. T. KEMP. 
ditions. This point was also noticed by Howlett (1, 224). I 
think the only way to obtain a reliable numerical determination 
of the plaques would be to prick the skin through a drop of 
osmic acid, examine a thin layer of this mixture, and count the 
relative number of plaques and red corpuscles. A separate 
determination would then have to be made by the usual methods 
for the red corpuscles, from which the number of plaques 
could be calculated. Even this method would not be free from 
error, for Hayem (26) has shown that enough plaques adhere to 
the edges of the wound to make an appreciable difference in the 
extravasated blood. 
Probably the most suggestive observations in support of the 
haematoblast theory were made by Hayem (4, 330), who has 
found the plaques in the vaso-formative cells of the mesentery of 
newly-born kittens, where the young red blood corpuscles are in 
process of development, while both Hayem and Pouchet (1, 97) 
have described intermediate stages between the plaques and the 
red corpuscles.1 
We cannot regard the haematoblast theory as proved, but we 
must acknowledge that the relationships between the plaques and 
the red corpuscles are most striking; and it seems as though this 
is the most plausible explanation afforded. 
§ 3. 
Micro Chemistry. 
In studying the raicro-cliemieal properties of the plaques, a 
combination of methods was used.2 
With water the plaques swell up into large bladder-like struc- 
tures, which consist of a very pale, hyaline sphere with granules 
collected at one point on its surface. The hyaline part is not 
easily distinguished in water without a very fine lens. By run- 
ning iodine under the slips, the bladders may be studied to much 
better advantage. If the blood be so completely washed away 
that no fibrin will form, the bladders will begin to disappear 
1 The observation of Hlava just mentioned may also be taken in support of 
Hayem and Pouchet, the colored bodies which he distinguishes from the plaques, 
being plaques containing haemoglobin, and on the way to form red corpuscles. 
2 Pp. 301-303. “NEW ELEMENT” OF THE BLOOD. 
309 
after a day, although at the end of three days some of them may 
still be found, especially the smaller ones. (Plate, Figs. 8-12.) 
Acetic acid (.1 per cent.) acts like water, only more power- 
fully. 
Concentrated acetic acid readily dissolves the red corpuscles 
and plaques. The white corpuscles are gradually dissolved, the 
nuclei resisting the action of the acid the longest. 
Concentrated solution of potassium hydrate quickly dissolves 
everything in the field except the granules of the coarsely 
granular leucocytes. 
Solution of magnesium sulphate (27 per cent.). If this solu- 
tion be used on the finger (p. 301), the plaques will remain 
without breaking down for a considerable length of time. 
Cold.—One of the best means of studying the breaking down 
of the plaques is to observe them in .75 per cent, sodium chloride 
solution or Bizzozero’s fluid at a reduced temperature—e. g. 6° 
C. This will retard the clotting so that the different stages may 
be conveniently observed. 
A very low temperature—e. g., —1 to —{—2.5° C.—will prevent 
coagulation. If a preparation be kept at this temperature for a 
few hours, and then the temperature be raised, the plaques will 
break down, but no fibrin will be formed. This is interesting, 
for, as Hayem (5, 708) has pointed out, there must be some deep- 
seated chemical change produced by the cold, which prevents 
the plaques from yielding that which appears to make them 
essential to the process of coagulation. 
Concentrated solution of mercuric chloride preserves the 
plaques, but causes a precipitate in the blood which renders its 
use objectionable. 
Of the different staining fluids, methyl violet, gentian violet, 
and strong fuchsin, stain the plaques readily and deeply. They 
should be used in very dilute solutions, especially the methyl 
violet. 
Iodine, when run under the slip, appears to stain the plaques, 
and makes them stand out very satisfactorily. It is particularly 
adapted for studying the bladder stage. The coloration from 
iodine is not, however, a true stain, for the last traces of it may 
be washed away by water even after the iodine has been allowed 
to act for thirty-six hours. Bismarck brown, magenta, Kleinen- 310 
GEO. T. KEMP. 
berg’s haematoxylin and aqueous haematoxylin, give very good, 
and for permanence the best, staining of the plaques, but a 
longer time is required than for methyl violet, gentian violet or 
fuchsin. Eosin will also stain them, but to get satisfactory 
results it is necessary to use a tolerably strong solution and 
allow it to stand from twelve to fourteen hours. Aniline blue- 
black, borax carmine, Frey’s carmine and picro-carmine, do not 
stain the plaques even in forty-eight hours. 
I also used a double staining fluid of carmine and indigo-car- 
mine, recommended by Shakespeare and Norris,1 which is claimed 
to stain green any cell wdiich contains haemoglobin. Hoping to 
find another method of determining whether the plaques con- 
tained haemoglobin, I tried it several times, but wdthout success. 
§4. 
Origin of the Plagues. 
The question now arises : Are the plaques normal and inde- 
pendent elements of the blood, or have they been derived from 
the other elements of the blood, either normally or by the 
methods of preparation ? 
This is a subject on which the greatest diversity of opinion 
exists, all of the following views having been held, and experi- 
mental evidence adduced to support them: 
1. They are haematoblasts, or young red corpuscles. 
2. They are derived from the red corpuscles. 
3. They are derived from the white corpuscles. 
4. They are nuclei floating free in the blood. 
5. They are fibrin. 
6. They are globulin depositions produced by cooling of the 
blood. 
7. They are independent elements. 
While the red and white corpuscles were the only recognized 
histological constituents of the blood, it was most natural for 
observers to refer the plaques and granular masses to one of these 
elements, without considering the possibility of the plaques being 
independent elements themselves. 
1 Amer. Jour. Med. Science, Jan., 1877. “NEW ELEMENT” OF THE BLOOD. 
311 
The theory that they are derived from the red corpuscles has 
received but little support. Boettscher (1), who believes the 
red corpuscles to be nucleated, thinks that the plaques are free 
nuclei from them. Ehrlich (1, 405) thinks that in anaemia the 
red corpuscles break down and give rise to the structures pictured 
by Riess. 
Laker (1, 192) has made this particular point the object of 
special study, and he comes to the conclusion that the plaques 
are not derived from the breaking down of the red corpuscles, 
for he has produced this artificially in various ways; but of all 
the fragments which he examined, he did not find one that he 
would have mistaken for a plaque. The same subject has been 
investigated by Schimmelbusch (2, 216), with similar results. 
The view that the plaques are derived from the leucocytes has 
met with much more general acceptance, and has been defended 
by observers of high authority. 
Riess says (1, 224) that they are destruction-products of the 
white corpuscles, produced by an insufficient nutrition of the 
blood. In another article (2, 696) he says that in pathological 
conditions, when the number of leucocytes is increased, we find, 
as a rule, that the number of plaques is increased also. 
Even more to the point than the observations of Riess are 
those of Halla (2, 376-378), who has shown that while the 
number of plaques and leucocytes respectively is not necessarily 
increased or diminished at the same time, yet in general the 
cases where the plaques are found in the greatest abundance are 
those in which the leucocytes are also most numerous, or closely 
following a condition where they had existed in great numbers. 
As a matter of fact, the very cases in which we find this 
increase of leucocytes and plaques are generally those in which 
the number of red corpuscles is diminished. This fact was 
noticed by Riess (2, 696), and especially by Hayem (11, 120), 
who says that his view is supported by the results of over fifty 
observations. Thus we see that the points made by Riess and 
Halla, will serve as well to support Hayem’s haematoblast theory, 
as the theory of the derivation of plaques from the leucocytes. 
Poucliet (2, 136), from similarity of staining, thinks the 
plaques are derived from the protoplasm of the leucocytes ; and 
Howlett, whose work was done after Hayem and Pouchet, but 312 
GEO. T. KEMP. 
about the same time as, and independently of, Bizzozero, is also of 
the opinion (1, 224) that the plaques are derived from the leuco- 
cytes, his reasons for this belief being : that they are absent from 
scrapings of lymphatic glands, spleen, and the medulla of bones 
in the dog; and that they resemble processes of leucocytes. 
From the latter statement we would infer that Ilowlett regards 
the plaques as derived from the body, rather than the nuclei of 
the leucocyte, although he mentions that they stain like nuclei 
with the aniline colors. 
Hayem (4, 49) at first appeared to advocate the view that the 
plaques were independent elements (elements particuliers, et a 
tous les d6gres de leur Evolution, parfaitement distincts des leu- 
cocytes). In a later article (9,198) he says: “ The haematoblasts 
arise in the lymph; they are found in the protoplasm of the 
white corpuscles, from which they are set free before entering 
the blood, except in certain pathological conditions.” Fie de- 
scribes leucocytes found in the lymphatic glands, in which ready- 
formed plaques may be recognized. These elements possess a 
nucleus more or less hidden by the plaques. IFe does not men- 
tion the multinuclear leucocytes to which Hlava attaches so 
much importance, and it is possible that he had these under 
observation, regarding their nuclei as plaques. Lowit (3, 127) 
also has seen structures in the leucocytes, in warmed peptone 
blood, which he took to be plaques. * 
He (3, 125) believes that he has also seen hyaline drops exude 
from leucocytes, swim off in the plasma, and become plaques. 
He gives a figure illustrating this process, which I reproduce 
(Plate, Fig. 13, c), and from which I infer that what he has really 
seen is a plaque which is adhering to the edge of a leucocyte, 
and which has become a small bladder.1 He has taken the 
hyaline part of the bladder for a plaque, while the granular part 
is lost against the leucocyte. Sometimes small bladders are 
formed, the hyaline part of which is little, if any, larger than a 
plaque (Plate, Figs. 11 and 12). He admits also, that this 
process takes place very quickly in blood drawn into salt solu- 
tions, and cannot always be seen, while it very rarely can be 
traced; which is well in accordance with such an inference. 
Halla thinks the plaques are derived from leucocytes which 
1 See page 308. “NEW ELEMENT” OF THE BLOOD. 
313 
break down in the circulation, because they are found to be most 
numerous at the same time, or immediately after the leucocytes 
are found in greatest numbers. He thinks they are nuclei of 
leucocytes because they resemble these structures more than the 
rest of the cell. 
The most ardent supporter of the theory that the plaques are 
nuclei of leucocytes is Hlava. Hlava (1), after a severe criticism 
of Bizzozero’s work, gives the results of some experiments of his 
own on plaques and the origin of fibrin, in which he is led to 
the conclusion that the plaques are nuclei of the leucocytes, 
which are set free when the latter break up, whether in the 
circulation or in extravasated blood. He adduces the following 
facts in support of his theory: 
First (1,403): He admits that he was unable to find any white 
corpuscle breaking down while in the circulation, but he has seen 
wdiite corpuscles filled with plate-like structures. “Such white 
corpuscles have the greatest similarity to conglomerations of the 
true plaques, and stand perhaps in close relation to them.” 
Second: He divides the leucocytes into several varieties, and 
says that one definite variety of these—the multinuclear (see 
Plate, Fig. 14, a, 5, c, g)—always breaks down much more readily 
than the others when the blood is shed. The total number of 
leucocytes is much smaller before coagulation than after, and the 
few which remain are nearly all of the uninuclear variety. 
Third: He has observed (1, 404) that we find more plaques 
when we examine blood in salt solutions than when it is drawn 
into osmic acid. He gives as the reason for this that the leuco- 
cytes are set by the osmic acid, and therefore cannot break down, 
while a destruction of the multinuclear leucocytes takes place in 
salt solutions, giving rise to what have been taken for groups of 
plaques. 
Fourth: He has also found (1, 408) that as the multinuclear 
leucocytes break down, their nuclei stain less deeply with 
gentian violet. The same may be noticed with regard to the 
breaking down of the plaques; moreover, the plaques and nuclei 
have about the same size and shape. 
Fifth (1,394): He quotes Rauschenbacli that fibrin formed 
from a necrosis of the leucocytes has the same micro-chemical 
properties as the plaques; and from this he draws the conclusion 314 
GEO. T. KEMP. 
that the micro-chemical properties of the plaques and dead 
leucocytes are the same. 
Sixth (1,409): He explains the presence of plaques in the 
blood while yet in the vessels by supposing the breaking down 
of the leucocytes to be a perfectly normal occurrence in the 
circulation. To support this view he refers to the work of 
Schmidt, who, with Jackowicki, Birk and others, has found that 
free ferment exists in the blood, and is increased in amount 
during fever. 
In regard to the first point that Hlava makes, I can say that I 
have seen the corpuscle which he describes (Plate, Fig. 14, a, b, c, g), 
but have always regarded it simply as a multinuclear leucocyte. 
It is certain that dilute acetic acid brings the small bodies in 
them into prominence, just as with a nucleus, while there is no 
such reaction as that shown when the plaques are treated with 
this reagent. Furthermore, I have not observed any biconcavity 
in the small bodies within the leucocytes, such as is seen in the 
plaques. 
As to the second observation of Hlava, I must say that my 
observations, in several instances where my attention was directed 
particularly to this point, do not go to confirm it. 
In the third point Hlava is correct in stating that we find 
fewer plaques in osmic acid than in dilute salt solutions, but I 
think the reason of this is, that in salt solutions the plaques 
adhere to the slide or cover-slip, while the other elements are 
drawn hither and thither, or even washed away entirely. Thus 
the plaques from the whole drop of blood will collect together on 
the slide or cover-slip, and appear very much more numerous in 
proportion to the other elements than they really are. This very 
fact is taken advantage of to get a great many plaques under 
observation in the same field. 
To the fourth point, I can only say, that in all the preparations 
of clot wrhich I have examined I have not seen half-a-dozen 
leucocytes broken down, but I have seen groups of plaques 
which presented exactly the appearance which we would expect 
to find in a leucocyte breaking down ; and I am only able to feel 
1 See p. 802. When the blood is mixed in proper proportions with strong salt 
solution, which prevents them from adhering, their number is found to be about 
the same as in osmic acid. “NEW ELEMENT” OF THE BLOOD. 
315 
confident that it was not a leucocyte from the fact that many of 
the latter could be distinguished, of all varieties, which showed 
no signs whatever of disintegration; and, more important still, I 
have followed the process of breaking down in masses the 
individual plaques of which I had seen come together, and these, 
when well broken down, would sometimes present the appearance 
so easily mistaken for a leucocyte. 
Hlava’s fifth argument will not hold good, because by necrosis 
the micro-chemical properties of the leucocytes may be changed 
entirely. Hlava himself regards coagulation as a necrosis of the 
white corpuscles; but the normal plaques and leucocytes do not 
have the same micro-chemical properties, as I shall show later. 
Hlava’s sixth point also cannot be said to be free from objec- 
tion. In the first place, the plaques are biconcave, while the 
nuclei of the leucocytes have never been made out to be so. 
They also differ in micro-chemical properties, the nuclei resisting 
the action of concentrated acetic acid longer, and also staining 
more readily than the plaques with certain reagents. They are 
not found in the blood of ovipara, nor in the blood of inverte- 
brates, although multinuclear leucocytes are found in the blood 
of these animals. The fact that ferment may exist in the blood, 
and even be increased in fever, does not prove that the leucocytes 
break down, for the plaques were present also, and it may have 
been derived from them. 
Of the older writers, Simon and Andral have mentioned 
structures which were possibly plaques, in connection, with, the 
formation of fibrin. (See historical section.) 
The chief supporter of the view that the plaques are deposi- 
tions of fibrin is Ranvier (1, 215 and 217), who comes to this 
conclusion from a few micro-chemical resemblances which exist 
between them, one of which is that they are both, stained with 
iodine. Lavdovsky (2, 65) and Hall a (1, 222) think that the 
plaques are morphologically allied to fibrin, from the intimate 
connection which exists between the granular masses and the 
fibrinous network. Ranvier also makes this a strong point. 
Now, I have shown (p. 309) that iodine does not give a true 
stain to the plaques, and the same may be said with regard to fibrin; 
but in preparations stained with Bismarck brown, or methyl 
violet especially, we have the remnants of the plaques deeply 316 
GEO. T. KEMP. 
stained, wliile the fibrinous threads radiating out from them are 
nearly colorless, and can be seen to run directly up to the mass, 
from which they can be more or less sharply distinguished. The 
same thing may be more clearly shown if haematoxylin be used 
instead of Bismarck brown or methyl violet. To do this, let the 
plaques and fibrin stain thoroughly, then remove the excess of 
haematoxylin by acid. The last traces of color will be removed 
from the fibrin while the plaques are still deeply stained. 
These methods of staining do not show, beyond doubt, micro- 
chemical differences between the plaques and fibrin, for in the 
case of haematoxylin it is only natural that the stain should be 
removed from such delicate structures as the fibrin-threads 
before it is from the granular masses. In the case of methyl 
violet or Bismarck brown, the difference in staining between the 
granular masses and the fibrin-threads seems to be too great to be 
accounted for by their difference in size, yet I have never been 
able to wash the stain entirely away from the fibrin. The action 
of distilled water, however, shows a true micro-chemical differ- 
ence between the plaques and fibrin. The plaques take on the 
bladder structure described, and gradually dissolve, while the 
fibrin remains unchanged. 
The theory that the plaques are precipitated globulins is held 
only by Lowit.1 This observer believes (3, pp. Ill, 114-117, 
124-125) that the plaques are globulin depositions which at the 
body-temperature are normally in solution in the plasma, or 
contained in the leucocytes, but which are precipitated from the 
one by cooling, or extruded from the other when the blood is 
drawn. He claims to have produced plaques by precipitating 
separately both fibrinogen and paraglobulin (3, 124 and 125). 
Lowit believes that the plaques do not exist in circulating 
blood, for the following reasons: 
First (3, 85-93): Their number is found to vary when drawn 
into salt solutions of different strengths. 
1 Unfortunately, the numbers of the Sitzungsberichte of the Vienna Academy 
which contain Lowit’s two first articles (1 and 3) were not in the University 
library at the time this article was written, and I was only able to find them 
after Schimmelbusch’s later article (2) came to hand. The work of Lowit is 
rather out of the usual line, and I hope, in a future publication, to enter more 
closely upon his investigations and results than I shall be able to do here. “NEW ELEMENT” OF THE BLOOD. 
317 
Second (3, 89): If blood be drawn into 25 per cent, salt 
solution, no plaques will be found. 
Third (3, 107): The plaques in the blood of peptonized dogs 
are not so numerous when drawn as after standing for some 
time. 
Fourth (3, 111): The plaques cannot exist in circulating 
blood because they dissolve in the plasma1 at the body-tempera- 
ture. 
Fifth (3, 113): Plaques do not exist in the blood of cold- 
blooded animals. The element described by Hayem, Bizzozero 
and others, is the representative of a stage in the development 
of the white corpuscles. 
With respect to the first reason, we should expect the number 
to vary in dilute salt solutions, on account of the adhesion of the 
plaques to the cover-slip.2 Additional evidence in favor of the 
assumption that the plaques are present in the circulation, is 
derived from the fact that if a large drop of strong salt solution: 
28 per cent. MgS04, 25 per cent. NaCl, etc., be put on the finger, 
and the finger pricked through the drop so as to get a small 
amount of blood in comparison to the drop of salt solution, the 
plaques are found isolated and floating freely, like the corpuscles, 
without showing any tendency to adhere. In this case the 
number of the plaques is about the same as when drawn into 
osmic acid. 
In repeating the experiment on which Lowit bases his second 
reason, I can only say that although in some preparations the 
number of plaques was very small, I have never found a prepa- 
ration in which they were entirely absent. 
I have never worked with peptonized dog’s blood, and there- 
fore have no suggestions to ofier, at present, in regard to the 
third point. 
The fourth reason will be referred to with the work of Schim- 
melbusch. 
The fifth point seems rather to be against the assumption that 
the plaques are globulin precipitations produced by cooling, for 
the blood of cold-blooded animals undoubtedly contains both 
paraglobulin and fibrinogen, as well as leucocytes. 
Lowit grants that Bizzozero and others may have found 
1 His work was done with peptonized blood. 
2 See p. 302. 318 
GEO. T. KEMP. 
plaques in the circulating blood, but says they were produced by 
cooling down the mesentery. He says if the mesentery had been 
surrounded by .75 per cent. HaCI solution at 37°-40° C., the 
plaques would not have been found. 
Schimmelbusch combats the views of Lowit, supporting his 
position mainly on two observations: 
1. In repeating Lowit’s experiment with normal plasma, he 
was unable to confirm Lowit’s observations with the peptonized 
plasma. His method of procedure was as follows (2, 212): 
A piece of mesentery was cut out with a red-hot scalpel, so as 
to prevent hemorrhage, and the plaques in the capillaries were 
found to be quite homogeneous. The piece of mesentery was 
then heated up to 40° C., but the plaques refused to dissolve. 
2. He has followed Bizzozero in examining the circulation in 
the mesentery of the rabbit, and has found the plaques circulating 
in the blood through the mesenteric vessels. In this experiment 
(2, 213-217) he has taken the precaution to meet Lowit’s objec- 
tion by keeping the mesentery of the animal immersed in a 
bath of .75 per cent, salt solution at the body-temperature. The 
mesentery is not drawn out until the animal is in position in the 
bath; so that at no time is there an opportunity for a precipitate 
to occur on account of cooling. 
The last view to be considered is that the plaques exist in the 
circulating blood as normal, independent elements. 
That they are not due to changes produced in other elements 
after the blood is drawn, is shown by pricking the finger through 
a drop of osmic acid, by which process all the elements of the 
blood are set immediately upon leaving the vessel. It is also 
pretty conclusive proof, that four competent observers—Bizzozero 
(7, 17), Lavdovsky (2, 64), Hlava (1, 403) and Schimmelbusch 
(1, 100) —have seen them circulating through the vessels in the 
mesentery, and Osier (1,2 and 3) has seen them in the unin- 
jured vessels of the connective tissue of young rats. Pouchet 
(2, 136), prior to Bizzozero, made observations which should be 
mentioned here as proof of the existence of free plaques in the 
circulation, although he regards them as derived from the 
leucocytes, and as being haematoblasts. His observations were 
made on the mesentery of rabbits. By compressing a small 
vessel he was able to cause stasis; by then observing the different “NEW ELEMENT” OF THE BLOOD. 
319 
leucocytes he found them with plaques adhering to their surface, 
as well as plaques free in the plasma. He concluded that the 
plaques had exuded from the leucocytes; but I think, from the 
light of more recent investigations, we may safely explain this as 
we did the observations of Lowit—viz.: that the plaques are 
normally free in the plasma, and that they are found adhering to 
the leucocytes by reason of their viscosity. 
There is no doubt as to the existence of the plaques in the 
blood, and though recognizing their close resemblances to other 
structures also found there, we have not as yet sufficient evidence 
to believe them to be other than an independent morphological 
element. 
This view is held by Max Schultze (1, 36), Osier (1, 144 and 
2, 530), Bizzozero (7, 17 and 3, 348), Laker (1, 193), Lavdovsky 
(2, 64), Halla (2, 378) and Schimmelbusch (1, 100) and (2, 217). 
§5. 
Histology of Coagulation with Especial Reference to the Part 
Played by the Plaques. 
In this section it is not my intention to dwell upon the different 
chemical theories of the coagulation of the blood; hut I shall 
take up the histology of the formation of fibrin, and its relation 
to the histological elements of the blood, the plaques in particular. 
The granular masses1 already described as a stage of the 
breaking down of the plaques, have been found by many observ- 
ers, and various opinions are held as to their origin.2 Schmidt 
(5, 356 ; 562-567) says they come from peculiar forms interme- 
diate between the leucocytes and the red corpuscles. He calls 
this intermediate form “ Rothe Kornerkugeln.” This opinion 
is also supported by Sernner (1), whose work was done in 
Schmidt’s laboratory, and published the following year. About 
the same time that Semner’s work came out, Schmidt (3, 528 
and 529) described the white corpuscles as breaking down, and 
giving the appearance of a mass of granules, which float off 
singly or in small groups in the surrounding plasma. This 
description differs from that given by Schmidt (3, 562) for the 
1 See page 295. 
2 The views of the older writers may be found in the historical section. 320 
GEO. T. KEMP. 
breaking down of the “ Rothe Kornerkugeln,” so tliat he evi- 
dently makes a distinction; but he thinks that both have to do 
with coagulation. 
The view of Schmidt and Semner is adopted by Slevogt (1) 
and Feiertag (1), pupils of Schmidt. 
Heyl (1, 35-38) has observed that in Bizzozero’s fluid the 
number of granular masses increases as the number of leucocytes 
decreases ; also that both granular masses and leucocytes stain 
with methyl violet. From this evidence, together with the 
assumption that the leucocytes break down during coagulation, 
he draws the conclusion that the granular masses are derived 
from leucocytes. 
Riess (1, 245) and Hlava (1, 404-409) both maintain that the 
granular masses are chiefly formed from the leucocytes, and 
claim to have found intermediate stages. 
Osier (2, 531) and Laker (1, 148) have each given special 
attention to this point, and both deny most emphatically that 
the granular masses are derived from leucocytes. 
Lowit (1, 281) has studied the coagulation of lymph under the 
microscope, and has found nothing that could be taken for 
leucocyte detritus. 
The most generally accepted view in regard to the granular 
masses is that they are formed by the plaques. This view is 
held by Max Schultze (1, 36 and 37), Osier (2, 530), Hayem 
(5, 675 and 704 and 16, 215), Leube (1, 654), Bizzozero (3, 352 
and 7, 18), Davidson (1, 1028), Laker (1, 183), Halla (1, 217), 
Lowit (1, 297), and Schimmelbusch (1, 101 and 2, 220). 
Hlava believes that both plaques and leucocytes form granular 
masses. He thinks that the plaques break down only because 
they are parts of white corpuscles, and has failed to observe that 
the plaques break down sooner than the other white elements. 
Lowit thinks that the granular mass proper is made up of 
plaques, but that leucocytes are often enclosed. 
I have succeeded, in the course of my investigations, in getting 
preparations in which the individual plaques were seen to unite 
and form the granular masses. By keeping these masses under 
observation, I was able, as they broke down, to follow the process 
continuously until after fibrin was formed. 
The results of my observations are given in the introductory 
section. “NEW ELEMENT” OF THE BLOOD. 
321 
In addition to this, experiments were made, in which speci- 
mens of blood were taken at different intervals after extravasa- 
tion and hardened in osmic acid, so as to get the different stages 
in the formation of fibrin. 
The results of my work all go to show that the breaking down 
of the plaques is intimately connected with the formation of 
fibrin. 
The granular masses formed by the plaques become centres 
from which the threads of fibrin radiate. The threads are also 
deposited freely in the field, and often as long, needle shaped 
bodies (Figs. 5, 6 and 7), but there is generally a thicker deposit 
of fibrin in the immediate vicinity of the granular masses, 
especially the large ones, than is noticeable elsewhere. 
The plaques, either before or after breaking down, are not 
morphologically identical with fibrin, so that they do not con- 
tribute as such to the formation of the fibrinous network; the 
remnants which are seen enclosed by the threads of fibrin are 
held there mechanically, and are not an essential part of the 
reticulum. 
That fibrin has a tendency to radiate from the granular masses 
as centres,-has been observed by Riess (1, 744), Ranvier (2, 93), 
Hayem (5, 704: and 705), Leube (1, 654:), Bizzozero (7,19 and 20 ; 
12,531),Davidson (1,1028),Lavdovsky(2,65) andHalla(2, 222). 
Ranvier, Bizzozero, Davidson and Halla think there is a direct 
connection between the fibrin-threads and the granular masses, 
while Max Schultze (1, 38) is of the opinion that the threads 
pass over or through the masses, but do not proceed from them.1 
Halla (1, 218) has noticed that the fibrin-threads are deposited 
most thickly in the neighborhood of large masses of plaques. 
Fano (2, 394:; 4, 211) says that sometimes the plaques form the 
centres of radiation for the fibrin, but more frequently the leuco- 
cytes are such centres. 
In many observations that I have made, I have found that 
when the plasma is very much diluted, but little fibrin is formed. 
Under these circumstances the field is comparatively clear, and it 
can then be seen that no fibrin proceeds from the leucocytes, but 
all comes from the plaques, or is deposited freely in the field 
(Plate, Fig. 6). When the clot is thicker, the fibrin is seen to 
1 From his figures one would judge the contrary. 322 
GEO. T. KEMP. 
radiate in many instances from large masses, of wliicli it is 
impossible to decide positively, from their appearance, whether 
they are plaques or leucocytes (Plate, Figs. 5 and 7). 
Sometimes a leucocyte will be covered more or less with 
plaques which can easily be distinguished at first, but which later 
would make it appear as if the leucocyte had broken down 
(Plate, Fig. 3). In some specimens of clot, I have seen leucocytes 
with only a few adherent plaques (Plate, Fig. 12). In such cases 
the fibrin has been seen to radiate typically from the plaques at 
the edge, wThile it bore no relation to the free margin of the 
leucocyte. In specimens hardened in osmic acid and stained 
with carmine this point can be made out very clearly. The 
leucocyte stains with the carmine, while the plaques and fibrin 
do not. 
Hayem (5, 720) and Schimmelbusch (2, 242) have each pointed 
out that fibrin may be formed elsewhere than around the plaques. 
Schimmelbuscb (1, 102; 2, 236 and 242) believes that the fibrin 
does not show any preference for one or another of the histo- 
logical elements of the blood. Bizzozero (4, 109) is of directly 
the opposite opinion, and states that fibrin is deposited around 
the plaques, and nowdiere else. 
When the formation of fibrin is abundant, there is a thick net- 
work of threads all over the field, and every spot is seen to be 
full of fibrin, whether it contains histological elements or not 
(Plate, Fig. 7); but when the formation of fibrin is scant, it will 
nearly always be noticed that the fibrin is deposited most thickly 
in the neighborhood of the masses of plaques (Plate I, Fig. 6). 
Even in preparations where the fibrin is sparingly formed, the 
threads are deposited elsewhere than around the granular masses, 
and occasionally, though rarely, I have found granular masses 
around which the fibrin did not appear to lie any more thickly 
than in the clear field. The fact that nearly always the fibrin is 
deposited most thickly around the granular masses, even radiating 
from them as centres, while interesting and significant, is not 
conclusive proof that the plaques are connected with coagulation ; 
for the same adhesive property of the plaques which makes them 
adhere to each other, may also cause the threads of fibrin to stick 
fast as they separate out from the medium around them. This 
seems all the more probable when we consider that the fibrin as 
well as the plaques is sticky and adheres to the glass. “NEW ELEMENT” OF THE BLOOD. 
323 
Preparations in which the clot is scanty are generally obtained 
from blood when diluted with a reagent which retards coagula- 
tion ; and the fact that in these preparations the fibrin is deposited 
most thickly in the vicinity of the masses of plaques maybe due 
to the plaques giving up something which produces or hastens 
the coagulation, and that in dilute solutions this substance is 
more plentiful in the neighborhood of the granular masses than 
elsewhere. More will be said of this later. 
In a recent publication Laker has taken the view that the 
fibrin-threads are folds of a membrane which he calls the 
primary fibrin membrane. Laker (2, 156) finds that if a very 
thin layer of blood be gotten under the cover-slip, at the end of 
fifteen minutes no network will be visible, but the fibrin will be 
formed as a transparent membrane without differentiation. If 
now the cover-slip be pressed or moved, folds may be produced 
in the membrane which he believes to be identical with fibrin : 
first (2, 155), because their histological properties are the same, 
and secondly (2, 156), because with defibrinated blood he failed 
to get the membrane. There is no relation between these folds 
of the membrane and the cellular elements, the latter being 
found just as they were caught and held by the viscous substance 
of which the membrane is formed. 
Reference has been made to such a formation of fibrin by 
Nasse,1 Anderson2 and Virchow.3 Rauschenbach (1, 8 and 9) 
and Halla (1, 232) describe a membranous form of fibrin from 
the coagulation of fibrinogen and paraglobulin with ferment. 
This membrane is transparent and homogeneous, and can be 
plainly seen only by sundry folds in it. 
The view of Laker is almost identical with that of Virchow,3 
who, as early as 1856, says: “ The fibrin-coagulum is at first 
a homogeneous, structureless mass in which the appearance 
of fibres is given only by foldings on its surface or by splitting 
and rolling up from the edges After the coagulation is 
complete we have a homogeneous jelly in which lighter cells are 
suspended Any shaking or jarring, or movement of one 
•Nasse, Das Blut, p. 40. Also Muller’s Archiv 1841.—Quoted from Holz- 
mann, Archiv fur Anatomie und Physiologie, Phys. Abth., 1885, p. 220. 
2 Froriep’s Notiz, 1844, p. 676.—Quoted from Holzmann, op. cit. 
3 Virchow, Gesammelte Abhandlung, 1856.—Quoted from Holzmann, op. cit. 324 
GEO. T. KEMP. 
part of the coagulum upon another, produces folds, which in a 
thick layer are confined to the surface, hut in a membrane involve 
the whole thickness. The folds sometimes form asters and 
sometimes a reticulum; at other times they are parallel, but 
they always appear as fine, smooth lines. . . . . The cellular 
elements are [simply mechanically] retained by the clot.” 
Rindfleisch and Hermann have also mentioned a homogeneous 
stage in the coagulation of the blood which precedes the formation 
of fibrin-threads. Rindfleisch (1, 159) describes numerous splits 
and cavities which are seen as the threads of fibrin are formed 
by consolidation (festwerden) from the homogeneous jelly, while 
Hermann (1, Yol. 1, p. 253) thinks the threads are formed by a 
process closely resembling crystallization. 
Schimmelbnsch (1, 102; 2, 236) insists very emphatically 
that the formation of threads of fibrin is a true crystallization 
process, and does not believe there is a previous stage, either 
homogeneous or granular. Ranvier (1, 21T) regards the plaques 
as fibrinous, and says that when the blood begins to clot, processes 
run out from them which form the groundwork (premieres trav6es) 
of the reticulum ; this then develops by depositions of fibrin, just 
as a crystal of copper sulphate will grow in a solution of that salt. 
Hayem (5, 719) does not believe in the fibrinous character of 
the plaques, but nevertheless compares the formation of fibrin to 
a “ sort of crystallization starting from small crystals already 
formed.” 
1 have often noticed the membrane which Laker has observed, 
and I have also noticed that folds of this membrane, when fine, 
cannot be distinguished from threads of fibrin ; but I do not think 
that all fibrin-threads can be regarded as folds of such a mem- 
brane. The fibrin-threads appear to be separated out from a 
homogeneous mass in which they lie, but when two threads are 
seen to cross, they can often be seen to be distinctly marked off 
from each other, the one lying beneath the other, and not con- 
nected as they would be if both were folds of the same mem- 
brane. Furthermore, formations of fibrin have often been 
obtained in which no traces of a membrane are to be found, if 
such ever existed. This is generally the case in specimens of 
blood considerably diluted. 
I think it most jprobable that what Laker describes as a mem- “NEW ELEMENT’ OF THE BLOOD. 
325 
~brane is a layer of the homogeneous substance described by 
Virchow, Rindfleisch and Hermann, which is essentially of the 
same composition as fibrin, and from which the fibrin-threads 
are formed by a process very closely resembling crystallization, if 
not identical with it. 
If a specimen of blood be drawn into a beaker and allowed to 
clot, it will first be seen to “ set ” into a firm jelly, in which the 
larger fibrin-cords develop later, and by their contraction squeeze 
out the serum. How, the consolidation (festwerden) of this 
colloid mass into threads, with the formation of splits and cavities 
(Spalten und Lucken) between them, as described by Kindfleisch, 
may be looked upon as “ a process closely resembling Crystalli- 
zation,” as Hermann defines it. The definite form which the 
fibrin-threads take, especially the typical needle-shaped form of 
the threads when deposited isolated in scanty clot, speaks strongly 
in favor of a crystallization ; while the subsequent toughening 
and contraction of the threads show a clear resemblance to the 
coagulation of certain proteids, notably myosin. 
In fact, it appears that we have in blood an interesting process 
which may be regarded as intermediate, in a certain sense, 
between a true crystallization on the one hand, and the coagula- 
tion of certain proteids, such as myosin, etc., on the other. 
From the foregoing, I think it is evident that there is no histo- 
logical connection between the plaques and fibrin, so that if the 
plaques are involved at all in coagulation, the connection 
must be a chemical one; that is, the plaques must give up some- 
thing to the plasma at the same time that they break down. 
Hayem (5, 704 and 705), Bizzozero (4, 101-103; 7, 19), 
Lavdovsky (2, 65), Halla (1, 222; 2, 378) and Ferraro (1, 295) 
have pointed out that fibrin is formed pari passu with the 
breaking down of the plagues ; and both Hayem and Bizzozero 
have shown that reagents or conditions which retard the breaking 
down of the plagues, retard to precisely the same extent the 
formation of fibrin; while reagents which preserve the 
PLAQUES PREVENT THE FORMATION OF FIBRIN ALTOGETHER. 
I think the strongest evidence we have of a connection between 
the plaques and the clotting of the blood is derived from these 
facts, and all my observations thus far completed confirm them 
in every respect. 326 
GEO. T. KEMP. 
Schimmelbusch (2, 238 and 239) notices the observations of 
Hayem and Bizzozero, that reagents which affect the breaking 
down of the plaques also affect coagulation. He does not attach 
much importance to this, but says that the reagents employed 
may possibly prevent coagulation, even if the plaques broke 
down. He admits, however, that there is more weight to be 
attached to the action of cold and warmth respectively slowing 
or hastening the breaking down of the plaques. 
Schimmelbusch (1, 102 and 103; 2, 242) does not believe that 
there is any connection between the plaques and coagulation. 
The chief objections which he urges against this assumption may 
be given briefly as follows : 
1. He refers to the experiment of Bizzozero (2, 308; 4, 103), 
in which this observer found that in the blood of animals after 
death the plaques retained their original form as long as the 
blood remained fluid, while in blood that had clotted the plaques 
were found more or less broken down. Schimmelbusch (2, 309) 
made several experiments in connection with this question, and 
was led to deny the parallelism between the breaking down of 
the plaques and the clotting of the blood. He opened the hearts 
of cats, dogs and man, at intervals of from one to twelve hours 
after death, and took a small bit of fibrin, which he transferred 
to .75 per cent. HaCl solution, or better, to osmic acid. Upon 
teasing out this fibrin, he found a great number of plaques, free 
and intact, lying on and between the threads of fibrin. On the 
other hand, he found that in the hepatic vein the blood was slow 
to clot, but the plaques changed rapidly. The plaques, however, 
were star-shaped, and less distorted than in arterial blood, owing 
to the absence of fibrin. From these observations he draws the 
conclusion that there is not an inseparable connection between 
the breaking down of the plaques and the coagulation of the 
blood. 
This work of Schimmelbusch is exceedingly interesting, as 
yielding positive results which tend to disprove the relation 
between the plaques and coagulation, so clearly indicated by the 
work of Hayem and Bizzozero. I do not think, however, that 
they can be regarded as conclusive proof to the contrary, as I 
hope presently to show. 
To Bizzozero’s observations that solutions of neutral salts “NEW ELEMENT” OF THE BLOOD. 
327 
would retard the breaking down of the plaques, and also the 
formation of fibrin, Schimmelbusch (2, 238) objects that even if 
plaques broke down, the salt might still prevent coagulation by 
acting chemically on what was set free. Now, it seems to me 
that this same objection maybe urged against the conclusiveness 
of Schimmelbusch’s own observations on the hepatic blood. We 
know that a number of ferments and other substances can be 
extracted from the liver after death, and it is not impossible that 
some of these may act in the manner described.1 
The fact that well-preserved plaques are found enclosed in fibrin 
taken from the heart some time after death, cannot be regarded 
as conclusive proof that the plaques are not connected with 
the formation of clot, unless we knew definitely what they were 
supposed to yield, and at the same time could be sure that the 
plaques are the only source from which the substance produced 
could be derived. We should also have to knojv positively that 
all the plaques were well preserved, and that none of them had 
broken down. 
A striking illustration of this came up in my own work a short 
time ago. In a series of transfusion experiments recently made 
in this laboratory by Professor Martin and Mr. J. P. Campbell, 
peptone was tried to render the blood incoagulable. It was 
found to work very well for dogs, but was without effect on cats. 
Thinking that this might throw some light on the subject of the 
relation of the plaques to coagulation, I took some specimens of 
peptonized cat’s blood for examination. The blood was drawn 
from the carotid artery, through a glass canula and a piece of 
rubber tube about six inches in length. Some specimens were 
taken and hardened in osmic acid with as little loss of time as 
possible. These were found to be the same as perfectly normal 
blood. What was my astonishment, however, upon examining 
specimens from blood which had stood about two minutes, to find 
a beautiful typical fibrin-reticulum, with hundreds of plaques 
well preserved and enclosed in the network!3 I naturally 
1 Haycraft thinks he has found such a ferment in the buccal glands of the 
medicinal leech, which prevents the action of fibrin-ferment. (See Proceedings 
of the Royal Society, London, Vol. 36, p. 478.) 
2 Preparations of peptone blood were obtained in the following manner: When 
it was desired to harden a specimen of blood as soon as possible after being drawn, 
one surface of a cover-slip was touched to the stream of blood as it flowed from 328 
GEO. T. KEMP. 
thought I had found a proof that the blood would clot without 
the plaques breaking down; but upon looking into the matter, I 
found that the tube was not clean, but that blood had been 
drawn through it shortly before I took my specimens, and that 
the walls were filled with pieces of adherent clot. Another cat 
was peptonized, and some specimens of blood drawn through a 
tube already used in the carotid, while others wTere drawn 
through a fresh tube in the femoral. 
The blood drawn through the dirty tube was found to clot a 
little more quickly than that drawn through the clean tube. In 
blood drawn through the clean tube, coagulation went hand in 
hand with the breaking down of the plaques. From blood taken 
through the dirty tube, specimens could be gotten in which there 
was a copious formation of fibrin, but in which nearly all the 
plaques were well preserved. I could always find some broken- 
down plaques, but whether these were brought from the clot on 
the sides of the tube, or whether they broke down as the blood 
flowed through, of course cannot be decided definitely. 
The difference in time between the clotting of blood drawn 
through the dirty tube, and that of the blood drawn through the 
clean tube, though slight, was noticeable, and this, together with 
the fact that when coagulation first sets in, the plaques are found 
preserved in the specimen from the dirty tube, while they are 
broken down in the specimen from the clean tube, would seem 
to show that in passing through the dirty tube the blood took up 
something which brought about coagulation before the plagues 
broke down, but which is also formed later by the plagues when 
they break down. 
From what we know of coagulation at the present time, it 
seems probable that the agent in fibrin-formation that would be 
most apt to conduct itself in this way is the ferment. 
Hayem (5, 720) and Bizzozero (2, 320) each think that the 
the tube ; a stream of .75 per cent. NaCI solution from a wash-bottle was then 
directed against it immediately, so as to wash off the excess of red corpuscles, 
and the slip was dropped into a watch-glass of 1 per cent, osmic acid. 
To obtain specimens of the blood after any desired interval of time, the blood 
was drawn into watch-glasses and a cover-slip laid on its surface (the cover-slip 
will float, so that only the under surface comes in contact). At the end of the 
desired interval the cover-slip was lifted off with a pair of forceps, washed, and 
hardened as above. “NEW ELEMENT” OF THE BLOOD. 
329 
part played by the plaques is to furnish something essential to 
the coagulation, and both agree that ferment is in all probability 
the agent in question. Bizzozero (2,319; 4,111; 6,278; 
13, 353) attempts to prove this by his well-known experiment of 
whipping blood with threads, by which operation he gets them 
full of adherent plaques and some leucocytes. The threads are 
carefully washed in .75 per cent. NaCl solution, and then added 
to a liquid coagulable with ferment. A clot is formed sooner or 
later, according as there are many or few plaques adhering to the 
threads (4, 111). To eliminate the objection that the adherent 
leucocytes had caused the clot, as well as to disprove the theory 
that the leucocytes yield fibrin-ferment, he took specimens of the 
same coagulable liquid, and to them he added pieces of spleen, 
lymphatic gland, red marrow, etc., all of which contain leuco- 
cytes, and got no coagulation ; from which he concludes that the 
leucocytes do not yield ferment, and therefore the ferment must 
have come from the plaques adhering to the threads. 
Rauschenbach (1, 76-95) enters into a detailed and lengthy 
criticism of Bizzozero’s work, in which he raises two objections 
to the experiment just described: First, that the threads may 
have absorbed ferment which was not washed out, and this it 
-was that produced coagulation; secondly, he says that the pro- 
toplastic fluid used by Bizzozero1 is only a test for free ferment. 
He thinks the ferment is combined in some way with the leuco- 
cytes, and that the MgSCb prevents it from being set free. 
The first objection is undoubtedly valid. By granting the 
second, we exclude the leucocytes adhering to the threads from 
a share in the formation of ferment, if any is formed by what 
does adhere to the threads. It then remains to be explained 
why the time of coagulation should depend upon the number of 
plaques, and the experiment remains of value as offering strong- 
support to the theory that ferment may be derived from the 
plaques. 
1 Bizzozero used a fluid recommended by Schmidt. It is made by drawing 
blood into a 28 per cent, solution of MgS04 in the proportion—blood, MgS04 
:: 1 : 3. The salt solution should be at 0°. When the blood is drawn into it, 
a constant, but not violent, stirring is kept up until the two are well mixed. 
The mixture is then allowed to stand until the cellular elements have settled. 
The plasma is then drawn off from the top layer and filtered at 0°, and this is 
then the protoplastic liquid. 330 
GEO. T. KEMP. 
The question of the formation of thrombosis is not without 
interest as bearing upon the question of coagulation. 
Before the plaques were well known, there was a great con- 
sensus of opinion among physiologists and pathologists that the 
“ white thrombus ” which is first formed around a lesion in the 
vascular wall was composed of leucocytes. Within the last four 
years this question has received the special attention of Osier 
(4; 2,531),Bizzozero (7,19; 3,358; 4,117), Hayem (22, 654 and 
655), Ferraro (1, 298) and Lubnitzky (1, 207), all of whom find 
that the white thrombus is not composed of leucocytes, as formerly 
supposed, but of plaques. 
In connection with my work on the plaques, I have taken up 
the subject of thrombosis, and hope to have my results ready for 
publication in the near future. I may say now, however, that 
my work thus far completed tends to confirm the results of the 
later observers just mentioned, and attribute the formation of 
the white thrombus to an agglomeration of plaques around a 
lesion in the vascular wall, or a foreign body introduced into the 
vessel. 
From the foregoing it would appear pretty evident that there 
is an intimate connection between the breaking down of the 
plaques and the coagulation of blood. On the other hand, some 
evidence has been adduced to show that coagulation may some- 
times take place without the plaques. 
Lymph will clot, and Fano and Lowit (1, 295) state that it 
contains no plaques. Hayem (9), on the contrary, maintains 
that the lymph does contain plaques, and, in fact, that the plaques 
in the blood have their origin in the lymph. I have not yet 
studied the coagulation of lymph, so that I cannot say at present 
whether the clotting of lymph is similar to that of blood, or 
whether differences exist that will throw light on the relation of 
the plaques to the coagulation of the blood. 
Lowit (1, 293) also claims to have obtained specimens of blood 
in 25 per cent. NaCI solution in which there were no plaques, 
but from which he could get a clot. I have repeated this experi- 
ment of Lowit, but with different results.1 
The strongest evidence yet brought forward against the suppo- 
sition that the plaques are necessarily connected with the coagu- 
1 See p. 317. “NEW ELEMENT” OF THE BLOOD. 
331 
lation of tlie blood, is derived from an experiment of Lowit 
(1, 297). This observer found that if rabbit’s blood be mixed 
with 28 per cent. MgSCh solution, and allowed to stand, the 
whole operation being conducted at 0°, the corpuscles will soon 
settle, and plasma from the top layer may be obtained in which 
there are numerous isolated plaques, a few red corpuscles, and no 
leucocytes. On adding wTater the plasma will not clot, although 
the plaques in it are numerous, whereas if fibrin ferment be 
added it soon forms a firm jelly. 
I have not had an opportunity, since consulting Lowit’s article, 
to repeat this experiment.1 Hayem (5, 708) has shown that 
plaques which have been subjected to temperatures of 0° 0. and 
under for a considerable length of time, lose their power to bring 
about coagulation when the temperature is raised. This I have 
found to be true, at least for a thin film of blood and .75 per cent. 
Nad solution, in an ordinary specimen for microscopic exami- 
nation. It may be that the action of MgSCh is somewhat similar, 
and produces deep-seated chemical changes in the plaques that 
destroy their power to yield the agent active in producing coagu- 
lation. 
From a careful consideration of both sides of the question, it 
is clear that much can be said for and against the theory that 
the plaques take part in the coagulation of the blood. But the 
fact that the breaking down of the plaques and coagulation go 
hand in hand in blood as nearly normal as it is possible to observe 
it, can hardly be without significance, and speaks strongly in 
favor of a connection between the plaques and coagulation as it 
normally takes place. The evidence on the other side is mostly 
derived from experiments involving the use of strong salt solu- 
tions and other very abnormal conditions, which detracts from 
its conclusiveness with regard to normal blood. 
Three other methods of formation of fibrin have been given 
by different observers, besides those already considered: 
1. From the red corpuscles. 
2. From the leucocytes, the leucocytes playing the part 
ascribed in this paper to the plaques. 
3. From the leucocytes by a process of necrosis. 
Each of these will be taken up in turn. 
1 See footnote, p. 316. 332 
GEO. T. KEMP. 
The work of Virchow,1 Hoppe-Seyler,2 Van der Horst,3 Heyn- 
sius,4 Mantegazza,5 Landois,0 Semner,7 Dogiel,8 Hart9 and Hayem 
(24, 62), all goes to show that fibrin may be obtained from the 
red corpuscles, and Dogiel thinks it possible that there is a rela- 
tionship between the stromas of the red corpuscles and fibrinogen. 
The methods employed in most of the above work place the 
corpuscles under conditions which do not exist in the ordinary 
process of coagulation; and, besides, it is now not claimed that 
red corpuscles are exclusively the source of fibrin, so that it is 
sufficient to mention this work as a matter of interest, but which 
need not be further discussed, as the red corpuscles do not enter 
the field as rivals of the plaques, in any theory of coagulation. 
There are two observations, however, which should be mentioned 
here, viz.: those of Landois and Hart. 
Landois10 and Hart10 claim to have observed the stroma of the 
red corpuscles changed directly into fibrin. This was likely due 
to currents, as Landois admits. In speaking of the red corpuscles 
he says: “ At first we can recognize the contour of individual 
corpuscles, but as soon as a current is set ujp11 in the surrounding 
liquid, the masses of stromas are carried hither and thither, by 
which process the adhering stromas are drawn out into tough 
threads, the cell-contour being at the same time lost.” He 
attributes the rapid clotting which follows the mixing of blood 
of animals of different species to this process. 
In the case of thrombi in the vessels, Landois says: “ When 
once the stroma-fibrin is formed, the plasma-fibrin can be 
deposited around it, as around a foreign body.” 
The next theory to be considered is, that the leucocytes play 
the part in coagulation which in this paper has been ascribed to 
the plaques. 
1 Virchow, Gesammelte Abhandlungen, 1856. 
2 Landois’ Lehrbuch der Physiologie des Menschen, 4th edition, 1885, p. 57. 
3 Over de Biwitachtige stoffen van het bloed, Leyden, 1868. 
4 Pfliiger’s Archiv, Vol. II, pp. 29-49 ; also, Vol. Ill, pp. 414-424. 
6 Centralblatt fur die Medicinischen Wissenschaften, 1868, p. 292. 
6 Centralblatt fiir die Medicinischen Wissenschaften, 1874, pp. 420-422. 
I Ueber die Faserstoffbildungim Amphibien undVogelblut, etc., Dorpat, 1874. 
8 Centralblatt fiir die Medicinischen Wissenschaften, 1875. 
9 Quarterly Journal Microscopical Science, 1882, p. 255. 
10 Op. cit. 
II The italics are mine. “NEW ELEMENT” OF THE BLOOD. 
333 
The belief that the leucocytes break down during coagulation, 
thus furnishing ferment, has always gone hand in. hand with 
Schmidt’s theory of the origin of fibrin from three fibrin factors; 
and as Schmidt’s theory for a long time met with almost universal 
acceptance, it is natural that the theory of the breaking down of 
the leucocytes should have gained a firm hold on the minds of 
physiologists and pathologists. It is therefore not strange that 
often in work on subjects involving the question of coagulation 
of the blood, the leucocytes have been said to break down, not 
because the attention of the observer was directed to this point, 
but because it was a generally accepted fact. This will account 
for a large amount of literature that may be quoted apparently 
in support of the theory of the breaking down of the leucocytes 
as part of coagulation. 
This question has been the subject of many researches, however, 
and has been approached in many different ways, with almost as 
many varying results. 
Schmidt, Hoffman and Heyl have found that there were fewer 
leucocytes in blood after coagulation than before, and from this 
draw the conclusion that the leucocytes break down during the 
process. Heyl says there is a loss of 70 per cent, of the leucocytes, 
while Schmidt estimates the loss at 90 per cent. 
These results were obtained for defibrinated blood, and the 
obvious objection has been raised by Bizzozero, Laker and others 
that whipping the blood would injure the leucocytes, and hence 
these figures would not hold good for normal clot. Heyl (1, 30) 
himself recognizes the fact that leucocytes are destroyed by 
whipping, for he shows that if blood be shaken instead of 
whipped, the loss of leucocytes is considerably smaller. He also 
points out that by shaking a solution containing fibrinogen, 
paraglobulin and fibrin-ferment, the time of coagulation is not 
affected, while if the same be tried with blood, it hastens coagu- 
lation. He therefore concludes that the leucocytes are broken 
down by shaking, and that the reduced time of clotting is due to 
this. 
These observations of Heyl show pretty clearly that numerical 
results obtained from blood which has been shaken or whipped 
for defibrination, will not hold good for blood under normal con- 
ditions. It is obvious that Heyl’s observation on the acceleration 334 
GEO. T. KEMP. 
of clotting by shaking does not prove that the leucocytes have 
anything to do with it, for he had the plaques and the leucocytes 
together in the blood, and from the extreme vulnerability of the 
plaques, we should expect them to be broken down by shaking or 
whipping. This objection may be raised to the work of Feiertag 
(1), who made comparative observations on the number of leuco- 
cytes and plaques before and after coagulation. To prove a con- 
nection between the leucocytes and coagulation, Hlava (1, 396 
and 397) refers to work done by Heyl (1, 47), who, following v. 
Samson Himmelstjerna and Hoffman, injected septic matter into 
the veins of animals, and studied the relation of the increased 
number of leucocytes to the yield of fibrin. It is quite evident 
that these results throw no light on the subject, as here also the 
plaques as well as the leucocytes were involved. Laker makes a 
good point also: that if 71 per cent, of the leucocytes really did 
break down during coagulation, some of them should be caught 
in the act. Schmidt (3, 528) says that he has found all stages in 
the breaking down of the leucocytes, but admits that he has 
never seen progressive changes in one individual, while Banschen- 
bach compares the breaking down of the leucocytes to an explo- 
sion (Der Yorgang verlauft .... in fast explosiver Weise.)—a 
view also adopted Groth (1, 69 and 70). 
Hlava (1, 410) says he has seen leucocytes arrested in different 
stages of disintegration, but like Schmidt does not claim to have 
observed changes in the same leucocyte. 
A view directly opposite to this is taken by Bizzozero (7, 20), 
Laker (1, 19S), Lavdovsky (2, 65) and Lowit (1, 277, 280-283, 
302). Bizzozero and Laker are of the opinion that the leucocytes 
undergo no change whatever during coagulation. Lavdovsky 
says, “ The breaking down of the leucocytes, if it occurs at all, is 
of no importance in clotting, and can scarcely be perceived.” 
Bizzozero, Laker and Lowit have all seen the leucocytes in 
blood after coagulation, and Bizzozero and Laker have kept 
leucocytes under observation for several days without finding any 
disintegration. My experience has been the same. Hlava 
(1, 408, 409, 416) and Bauschenbach (1, 12 and 13) have 
advanced a theory, based on their direct observations, which will 
account for this—viz. : that there are different varieties of 
leucocytes, and that one variety breaks down readily, while others 
do not. “NEW ELEMENT” OF THE BLOOD. 
335 
My observations on this point have already been given (pp. 313, 
314), and, as is seen, do not confirm it. The results of my work on 
this subject receive confirmation from the work of Lowit on the 
lymph. Lowit has counted the leucocytes in a given field before 
and after coagulation, and finds their number to be the same. 
Lowit thinks that the leucocytes participate in coagulation by 
emitting something which dissolves in the plasma; but from 
work both on blood and lymph, he comes to the conclusion that 
they do not break dowm. 
This view, according to Bizzozero (4, 99), was first held by 
Mantegazza (1871), and I have seen no mention of any other 
advocate of it up to the time of Lowit. 
The following observations of Lowit have led him to the belief 
that the leucocytes take part in coagulation : 
He has found (1, 290) that if lymph containing leucocytes be 
added to hydrocele or ascites-fluid, it will cause coagulation. If, 
however, the lymph had been previously filtered through glass 
W'ool, and the leucocytes thus removed without breaking down, 
no clot would result from the addition of the lymph-plasma. 
He claims (1, 293) to get blood in which there are no plaques 
by drawing it into 25 per cent. NaCl solution. On diluting this 
mixture it will clot if leucocytes be present, otherwise not. 
He has observed (1, 301) that if blood be drawn into vessels 
surrounded with ice, and the different histological elements 
allowed to settle, they will be found in layers, some of which are 
particularly rich in one kind of corpuscles, some in another. If 
now a portion be taken from the layer which is richest in leu- 
cocytes, it will be found to produce coagulation more readily 
than specimens from layers poorer in leucocytes. These results 
are very striking, and would almost carry conviction on their 
face, but Lowit also mentions (1, 300) that the layer in which 
leucocytes are most plentiful also contains many masses of plaques, 
which, of course, detracts from the decisiveness of the experiment. 
Rauschenbach has succeeded in extracting ferment from leu- 
cocytes obtained from a number of sources, but his methods 
involved rubbing them up into a pulp, treating them with 
distilled water, salt solutions, the serum of blood from species 
other than their own, etc. These results are all interesting and 
suggestive, but do not show that ferment is derived from the 336 
GEO. T. KEMP. 
leucocytes as a normal part of the coagulation. It is well known 
that peptone injected into the veins of some animals prevents 
the blood from clotting. Such blood has been studied for infor- 
mation regarding coagulation chiefly by Fano and Wooldridge, 
and many points of great interest are suggested by their work. 
I may sav, however, that from this work there has been derived 
no positive proof that the leucocytes furnish ferment in ordinary 
coagulation. The ferment has always been obtained by the 
action of distilled water, a continued stream of carbon dioxide, 
etc.—conditions which are not present in the normal clotting of 
the blood. 
There is great need of more thorough microscopic investigation 
combined with the chemical; and as I propose, in continuing my 
work on coagulation, to take up a series of investigations in this 
field, I shall leave a detailed discussion of this subject for a future 
publication. 
We are now in a position to decide between the relative claims 
of the plaques and of the leucocytes to the function of yielding 
ferment, or being otherwise concerned in the coagulation of the 
blood. It can by no means be said that it is impossible to get 
ferment from the leucocytes. The results of many observers are 
strongly in favor of the assumption that, under certain circum- 
stances, the ferment can be so derived. Rauschenbach (1), 
Wooldridge (6, 417), Lea Green1 and Holzmann,2 claim to have 
extracted ferment from sources other than the cellular elements 
of the blood. Bizzozero (2, 321), while claiming for the plaques 
the power of furnishing ferment, expressly states that he does 
not attribute the origin of ferment exclusively to these elements. 
The point which I wish to emphasize is this: that whatever may 
occur in leucocytes subjected to abnormal conditions—from 
chemical reagents, crushing, concussion, from whipping of the 
blood, etc.—in the coagulation of the blood as it normally occurs 
there is no histological evidence that the leucocytes take part, 
while the contrary is true of the plaques. 
I may here add that I have seen apparently abnormal leuco- 
cytes in many specimens of clot. I do not, however, regard this 
as proof that the leucocytes break down during coagulation, but 
1 Journal of Physiology, Yol. IV, pp. 380-386. 
2 Holzmann, op. cit. “NEW ELEMENT” OF THE BLOOD. 
337 
tliink it more likely that the leucocyte was in that condition 
when the blood was drawn, or that, possibly, alterations were 
produced by its being caught in the contracting fibrin-threads. 
In specimens where the coagulation was studied in Bizzozero’s 
fluid, or at a reduced temperature, where the process is so re- 
tarded that its individual steps can be distinguished, I have 
never seen anything that could be taken for progressive changes 
in the leucocytes, while such changes in the plaques are most 
striking. 
The last theory to be considered is that fibrin is formed by 
a necrosis of the white corpuscles. 
This view is foreshadowed in Beale (1 and 2), and was next 
taken up by Weigert, whose object was to point out the interest- 
ing generalization that a necrosis of structured cells produced a 
substance resembling fibrin. He would gladly include blood- 
fibrin in this generalization, deriving it from a necrosis of the 
leucocytes, but he is forced to admit (2, 92) that “ Fibringerin- 
nung”is “ganz isolirt,” and to speak (1, 469) of the substance 
derived from the leucocytes as “ fibrinahnlich.” 
The later supporters of this view are Wooldridge and Hlava. 
Wooldridge believes that in ordinary coagulation of the blood 
there are two processes, one of which involves substances in 
solution, the other the cell-substance of the leucocytes. His view 
is briefly given by himself (2,418) as follows: 
“ There are two essential processes in the coagulation of the 
blood, one of which has been, hitherto, entirely wrongly appre- 
ciated or overlooked. This latter process is that the ‘ dead ’ 
plasma converts the white corpuscles directly into fibrin. At 
the same time, however, that this occurs, a substance is liberated 
from the cells which converts the fibrinogen also into fibrin. 
This is the other process. The substance which is liberated from 
the cells is fibrin-ferment.” 
The view of Hlava (1,414) is very similar to that of Wool- 
dridge. He believes that fibrin is produced directly from a 
necrosis of the leucocytes. For convenience of description he 
divides the formation of fibrin into four stages: 
1. The agglomeration of the leucocytes. 
2. The breaking down of these (the protoplasm breaking down 
before the nuclei). 338 
GEO. T. KEMP. 
3. The death of the nuclei (which according to Hlava are 
plaques). 
4. The solidification of the fibrin. 
At the third stage the ferment is set free, which causes the 
coagulation of the fibrinogen and paraglobulin. The fibrin is 
first granular, then fibrillar. 
I think the most likely explanation of these results has already 
been given—viz.: that groups of plaques may have been mis- 
taken for leucocytes, and that in many experiments he used a 
concentrated solution of mercuric chloride, which produces a 
precipitate of proteids in the plasma that obscures the whole 
field and vitiates the accuracy of any observations. 
Kauschenbach (1,52) has given attention to the formation of 
fibrin in the presence of many leucocytes, and his conclusion is 
that the fibrin formed by adding different kinds of leucocytes to 
plasma is true fibrin produced by the spontaneous coagulation of 
the plasma, the apparent differences being due only to the 
detritus of the leucocytes, and other foreign matter mechanically 
retained. 
SUMMARY. 
The conclusions arrived at in this paper may be given briefly, 
in summary, as follows : 
1. In addition to the red corpuscles and leucocytes, the blood 
normally contains a third histological element, the plaques. 
2. Although strong resemblances exist between the plaques 
and other histological elements of the blood, there is not yet 
sufficient evidence to establish a genetic connection. We are 
therefore obliged, for the present at least, to regard the plaques 
as independent elements. 
3. When the blood is drawn, the plaques break down almost 
immediately. This is not true of any other element of the blood. 
4. The breaking down of the plaques is intimately connected, 
in its time-relations at least, with the clotting of the blood. 
5. The connection between the breaking down of the plaques 
and the coagulation of the blood is not histological, but chemical— 
i. e., the plaques appear to give up a soluble substance which is 
active in coagulation. 
6. The active agent in question is most probably fibrin-ferment. “NEW ELEMENT” OF THE BLOOD. 
339 
7. Fibrin is deposited histologically independent of any of the 
cellular elements of the blood. 
8. When the clot is very scant, fibrin is deposited as long, 
needle-shaped, crystal-like bodies. 
In conclusion, I would express my thanks to Prof. Martin for 
encouragement and valuable suggestions throughout my work, 
and also to Dr. J. R. Duggan, who kindly assisted me in photo- 
graphing the plaques, and to whose skill and experience in 
micro-photography I feel my success in this direction has largely 
been due. 
BIBLIOGRAPHY. 
As the subject treated in this paper is comparatively recent, and as 
no attempt has yet been made to collect the bibliography, which is 
now becoming voluminous, I have been advised to publish the bibli- 
ography as here appended. 
The numbering of the articles does not in every case correspond 
with the order in which they appeared, but as an attempt to rearrange 
the numbers so as to get the articles in the proper order of publication 
would involve not only considerable labor, but also a great risk of 
getting the references confused, I have thought it advisable not to 
make any change. 
Afanasev.—1, 1884. Deutsches Archiv fiir klinische Medicin, 
Band 33, p. 217. Ueber den dritten Formbestandtheil im normalen 
und pathologischen Zustande, und iiber die Beziehung desselben 
zur Regeneration des Blutes. 
2, 1884. Vrach (The Physician), St. Petersburg, 1884, pp. 270, 
284, 304, 321. (Same as 1, but in Russian.) 
Beale.—1,1864. Transactions of the Royal Microscopical Society, 
1864, p. 32. Observations upon the nature of the red blood corpuscle. 
2, 1864. Ditto, p. 47. On the germinal matter of the blood, 
with remarks upon the formation of fibrin. 
Bizzozero.—1, 1881. R. Accademia di Medicina di Torino, 
Dec., 1881, pp. 124-128. Di un nuovo elemento morfologico del 
sangue dei mamiferi e della sua importanza nella trombosi e nella 
coagulazione. 
2, 1882. Virchow’s Archiv, Vol. XC, pp. 261-332 (with plate). 
Ueber einen neuen Formbestandtheil des Blutes und dessen Rolle 
bei der Thrombose und der Blutgerinnung. 340 
GEO. T. KEMP. 
3, 1882. Archives Italiennes de Biologie, Vol. II, pp. 345-362 
(with plate). D’un nouvel Element morphologique du sang, et de son 
importance dans la thrombose et dans la coagulation. 
4, 1883. Archives Italiennes de Biologie, Vol. Ill, pp. 94-121. 
(Continuation and conclusion of 3.) 
5, 1882. Archives Italiennes de Biol., Vol. I, pp. 274-276. Sur 
les petites plaques du sang des mammiferes. 
6, 1882. Arch. Ital. de Biol., Vol. I, pp. 276-278. Les petites 
plaques du sang et la coagulation. 
7, 1882. Centralblatt fur die medicinischen Wissenschaften, 1882, 
pp. 17-20. Ueber einen neuen Eormbestandtheil des Saugethier- 
blutes und die Bedeutung desselben fur die Thrombosis und Blut- 
gerinnung uberhaupt. 
8, 1882. Centralb. f. d. med. Wiss., 1882, pp. 161-163. Die 
Blutplattchen der Saugethiere und die “Invisible Corpuscles” von 
Norris. 
9, 1882. Arch. Ital. de Biol., Vol. I, pp. 1-4. Sur un nouvel 
element morphologique du sang chez les mammiferes et sur son 
importance dans la thrombose et dans la coagulation. 
10, 1882. Centralb. f. d. med. Wiss., 1882, pp. 563-564. Blut- 
plattchen und Thrombosis. 
11, 1882. Editorial in Lancet, 1882, Vol. I, p. Ill, on Bizzozero’s 
work. 
12, 1883. Centralb. f. d. med. Wiss., 1883, pp. 529-532. Die 
Blutplattchen im peptonizirten Blut. 
13, 1882. Centralb. f. d. med. Wiss., 1882, pp. 352-355. Blut- 
plattchen und Blutgerinnung. 
Boettscher.—1, 1877. Virchow’s Archiv, Vol. LXIX, pp. 295- 
307 (with Plate XI). Ueber einige Veranderungen welche die rothen 
Blutkorperchen in Extravasaten erleiden. 
Davidson.—1, 1882. Lancet, 1882, Vol. I, pp. 1028-1029. A 
clinical study of the small granular cells of the blood. 
Donne.—1, 1844. Cours de microscopie, etc. Paris, 1844. (Atlas 
bound separately.) 
2, 1842. Compt. Kend. de l’Acad. des Sciences, Vol. XIV, pp. 
366-368. 
Ehrlich.—1, 1880. Berliner klinische Wochenschrift, 1880, p. 
405. Proceedings of the “ Gesellschaft der charite Aerzte ” in Berlin. 
Eano.—1, 1882. Archivo per le Scienze Mediche, Vol. V, pp. 
116-161. 11 peptone ed il tripton nel sangue, e nella linfa. 
2, 1882. Ditto, pp. 392-395. Contribuzione alio studio della 
coagulazione del sangue. “NEW ELEMENT” OF THE BLOOD. 
341 
3, 1882. Archives Italiennes de Biologie, Yol. II, pp. 146-154. 
De la substance qui empeche la coagulation du sang et de la lymplie 
lorsqu’ils contiennent de la peptone. 
4, 1882. Centralblatt fur die medicinischen Wissenschaften, pp. 
210-211. Beitrage zur Kentniss der Blutgerinnung. (Almost the 
same as 2.) 
5, 1882. Lo Sperimentale, Yol. XLIX, p. 270. (Same title as 2.) 
6, 1881. Arcbiv fur Anatomie und Physiologie, 1881, Physiol. 
Abtheilung, pp. 277-296. Das Yerhalten des Peptons und Tryptons 
gegen Blut und Lymphe. 
7, 1882. Lo Sperimentale, Yol. XLIX, p. 459. Della sostanza 
che impedische la coagulazione del sangue e della linfa peptonizzati. 
Feiertag.—1,1888. Beobachtungen iiber die sogenannten Blut- 
plattchen (Blutscheibchen). Inaug. Diss., Dorpat, 1883. 
Ferraro.—1, 1884. Arch. Ital. de Biologie, Yol. 5, pp. 293-315. 
Experiences et nouveau resultats sur la physio-pathologie des arteres, 
etc. 
Feuerstack.—1,1883. Zeitschrift fur wissenschaftliche Zoologie, 
Yol. 38, pp. 134-163 (with six woodcuts). Die Entwickelung der 
rothen Blutkorperchen. 
Groth.—1,1884. Ueber die Schicksale der farblosen Elemente im 
kreisenden Blute. Inaug. Diss., Dorpat, 1884. 
Halla.—1,1883. Zeitschrift fur Heilkunde, Prag, 1883, pp. 198- 
251. tjber den Haemoglobingehalt des Blutes, und die quantitativen 
Yerhaltnisse der rothen und weissen Blutkbrperchen bei acuten 
fieberhaften Krankheiten. 
2,1883. Ditto, pp. 331-379. Continuation of 1. 
Hart, Mrs. Ernest.—1,1882. Quarterly Journal of Microscopical 
Science, 1885, p. 255. A note on the formation of fibrin. 
Hayem.1—1,1877. Comptes Rendus de l’Academie des Sciences, 
Paris, Vol. LXXXIV. 
1. Pp. 1166-1169. Des caract&res anatomiques du sang chez le 
nouveau ne, pendant les premiers jours de la vie. 
2. Pp. 1239-1242. Sur lanatureet la signification des petits globules 
rouges du sang. 
2, 1877. Compt. Rend, de l’Acad. des Sc., Vol. LXXXY. 
1 In addition to the papers here given, the Comptes Rendus de la Societe de 
Biologie contains a good many communications by Hayem, but as these are 
essentially contained in his articles in the Gazette Medicate and the Comptes 
de PAcademie des Sciences, I have not taken the unnecessary trouble to hunt 
them up. 342 
GEO. T. KEMP. 
1. Pp. 907-909. Sur revolution des globules rouges dans le sang 
des vertebres ovipares. 
2. Pp. 1285-1288. Sur revolution des globules rouges dans le sang 
des animaux superieurs (vertebres vivipares). 
3, 1878. Compt. Rend, de l’Acad. d. Sc., Vol. LXXXVI, p. 58. 
Sur la formation de la fibrine du sang, etudiee an microscope. 
4, 1878. Gazette Medicale de Paris, 1878. 
1. Pp. 15-17. Note sur les caracteres et Involution des hemato- 
blastes chez les ovipares. 
2. Pp. 43-45. Continuation of 1. 
3. P. 49. No title. Abstract of HayenPs remarks on Pouchet’s 
theory of the formation of red corpuscles before the Societe de Biologie. 
4. Pp. 60-61. Same title as 2, (2). 
5. P. 107. Same title as 3. 
6. Pp. 121-122. Considerations nouvelles relatives aux elements 
du sang. 
7. Pp. 257-258. Note sur le sang du chat nouveau ne. 
8. P. 330. Communication sur la formation des globules rouges 
dans les cellules vasoformatives. 
5, 1878. Archives de Physiologie 1878, pp. 692-734 (with 2 
plates). Recherches sur revolution des hematies dans le sang de 
rhomme et des vertebres. 
6, 1878. Revue internationale des Sciences, March, 1878. Not 
accessible. Could not find title. 
7, 1879. Arch. d. Physiol., 1879, pp. 201-261 (with plate). Con- 
tinuation of 5. 
8, 1879. Ditto, pp. 577-613 (with plate). Continuation and 
conclusion of 5 and 7. 
9, 1879. Gazette Med. de. Paris, 1879, p. 198. Sur Porigin des 
hematoblastes. 
10, 1879. Ditto, p. 565. Sur le stroma des globules rouges. 
11, 1880. Gaz. Med., 1880. 
1. Pp. 119-120. Sur les caracteres anatomiques du sang particu- 
liers aux anemies intenses et extremes. 
2. Pp. 191-192. Sur les caracteres anatomiques du sang dans les 
phlegmasies. 
3. Pp. 215-216. Continuation of 11, (2). Same as 16. 
12, 1880. Compt. Rend, de l’Acad. des Sc., Yol. XC,. p. 225-228. 
Identical with 11, (1). 
13, 1880. Ditto, pp. 614-617. Identical with 11, (2). 
14, 1880. Ditto, pp. 708-711. Continuation of 13. Identical with 
11, (3). “NEW ELEMENT” OF THE BLOOD. 
343 
15, 1878. Recherches sur l’anatomie normale et pathologique du 
sang. Monograph. 144 pages. Masson Ed., Paris, 1878. 
16, 1880. Gaz. Med., 1880, pp. 215-216. See 11, (3). 
17, 1881. Gaz. Med., 1881, pp. 479-481. Contribution & l’etude 
de la structure des hematoblastes et des hematies. 
18, 1883. Arch, de Physiol, 3ieme Serie, 1883, Yol. I. 
1. Pp. 214-223. Contributions a l’etude des alterations morpho- 
logiques des globules rouges. 
2. Pp. 363-373. Des globules rouges a noyau dans le sang de 
l’adulte. 
19, 1883. Gazette Medicale, 1883, pp. 432-433. Note sur les 
“Plaquettes du Sang” de M. Bizzozero, et sur la “Troisieme Cor- 
puscule du Sang” ou “Invisible Corpuscle” of M. Norris. 
20,1883. Gazette hebdomadaire de Medecineet de Chirurgie, 1883, 
pp. 432-433. Same as 19. 
21, 1883. Ditto, pp. 856-859. Du role des hematoblastes dans le 
coagulation du sang. 
22, 1883. Compt. Rend. de l’Acad. des Sc., Yol. XCVI, pp. 653— 
655. Experiences demontrant que les concretions sanguines formees 
au niveau d’un point les6 des vaisseaux, deboutent par un dep6t 
d’hematoblastes. 
23- Compt. Rend. d. l’Acad. des Sc., Vol. XCII, pp. 82-92. Sur 
l’application de l’examen anatomique du sang au diagnostique des 
maladies. 
24, 1882. Revue des Sciences Medicates, Yol. XIX, pp. 61-62. 
Du proces de coagulation et de ses modifications dans les maladies. 
25, 1882. Compt. Rend, de l’Acad. des Sc., Yol. XCIY, pp. 200- 
202. De la crise hematique dans le maladies aigues a deervescence 
brusque. 
26, 1882. Compt. Rend, de l’Acad. des Sc., Yol. XCV, pp. 18-21. 
Sur le mecanisme de l’arret des hemorrhagies. 
27, 1883. Compt. Rend, de l’Acad. des Sc., Yol. XCVII, pp. 144- 
147. Nouvelle contribution 5, l’etude des concretions sanguins intra- 
vasculaires. 
28, 1883. Compt. Rend, de l’Acad. des Sc., Vol. XCVII, pp. 458- 
461. Same title as 19. 
29, 1883. Arch, de Phys, 3ieme Serie, 1883, Vol. II, pp. 247-256. 
De la crise hematique dans la fievre intermittente. 
Hermann-.—1. Handbuch der Physiologie. Leipzig, Edition 
1880. 
Heyl.—1, 1882. Zahlungsresultate betreffend die farblosen und 
die rothen Blutkorperchen. Inaug. Diss., Dorpat, 1882. 344 
GEO. T. KEMP. 
Hlava.—1, 1883. Archiv fur experimentelle Pathologie und 
Pharmakologie, Yol. XVII, pp. 392-418 (with plate). Die Beziehung 
der Blutpliittchen Bizzozero’s zur Blutgerinnung und Thrombose. 
2.1883. Fortschritte der Medicin, 1883, pp. 341-343. ZurHisto- 
genese des Fibrins. 
Hoffman.—1881. Ein Beitrag zur Physiologie und Pathologie 
der farblosen Blutkorperchen. Inaug. Diss., Dorpat, 1881. 
Howell.—1, 1884. Science, Vol. Ill, p. 46. The new morpho- 
logical element in the blood. (Review of the work of Hayem, 
Bizzozero and Laker up to date.) 
Howlett.—1, 1882. Lancet, 1882, Yol. I, pp. 223-224. On the 
granular matter of the blood. 
Laker.—1, 1883. Sitzungsberichte der kaiserlichen Akademie der 
Wissenschaften. Wien. Dritte Abtheilung, Vol. LXXXVI, pp. 173- 
201. Studien liber die Blutscheibchen und den angeblichen Zerfall 
der weissen Blutkorperchen bei der Blutgerinnung. 
2.—Ditto, Yol. XO, pp. 147-159. Die ersten Gerinnungser- 
scheinungen des Siiugethier Blutes unter dem Microscope. 
Landois.—1, 1874. Centralbl. f. d. med. Wiss., 1874, pp. 420- 
422. Microscopische Beobachtung der Fibrinbildung aus den rothen 
Blutkorperchen; Verhalten des Fibrins in der Blutbahn. 
2. Lehrbuch der Physiologie des Menschen. 4th Edition, 1885. 
Laptschinsky.—1, 1874. Centralbl. f. d. med. Wiss., 1874, pp. 
657-661. Zur Pathologie des Blutes. 
Lavdovsky.—1, 1883. Vrach (Russian for Physician), 1883, 
Nos. 11-15 (with woodcuts). On the question of the third element 
of the blood in man and some animals. (English translation of 
Russian title.) 
2- Synopsis of 1 in Hoffman and Schwalbe’s Jahresbericht, Vol. 
XII, pp. 64-65. 
Leube.—1, 1879. Berliner klinische Wochenschrift, 1879, pp. 
653-655 (with woodcut). Ein Fall von essentieller An'amie mit 
ubermassiger Entwickelung der Kornerbildungen im Blute. 
Lowit.—1, 1884. Sitzungsberichte d. kais. Akad., etc., Wien, 
Yol. LXXXIX, pp. 270-307. Beitragezur Lehre von der Blutgerin- 
nung. Erst Mittheilung. Ueber das coagulative Vermogen der Blut- 
plattchen. 
2, 1885. Fortschritte der Medicin, 1885, pp. 173-178. Die Blut- 
plattchen und die Blutgerinnung. 
3.1884. Sitzungsberichte der kais. Akad., etc. Wien, Yol.XC, pp. 
80-132. Beitrlige zur Lehre der Blutgerinnung. Zweite Mittheilung. 
Ueber die Bedeutung der Blutplattchen. “NEW ELEMENT” OF THE BLOOD. 
345 
4, 1885. Fortschritte der Medicin, 1885, pp. 276-278. Berichti- 
gung, die Blutpl'attchen betreffend. 
Lubhitzky.—1, 1885. Archiv f. exp. Path, und Pharm., Yol. 
XIX, pp. 185-209 (with 2 plates). Die Zusammensetzung des 
Thrombus in Arterienwunden in den ersten funf Tagen. 
Mayet.—1, 1882. Archives de Physiologie, 2d Series, Yol. IX, 
pp. 237-277. Recherches sur les alterations spontanees des elements 
colores du sang conserves dans le plasma l’abri de l’air. 
Nedsvetzki.—1, 1873. Centralbl. f. d. med. Wiss., 1873, pp. 
147-150. Zur Histologie des Menschenblutes. 
Norris.—1,1879. Transactions of the Birmingham Philosophical 
Society, 1879. 
2, 1882. Lancet, 1882, Vol. I, p. 163. The new blood corpuscle. 
3. Lancet, 1882, Vol. I, pp. 561-562. On the' claim of Prof. 
Bizzozero to the discovery of the fibrin-forming corpuscle. 
Osler.—1, 1874. Monthly Microscopical Journal, London, 1874, 
pp. 141-148. An account of certain organisms occurring in the 
liquor sanguinis. 
2, 1882. Centralbl. f. d. med. Wiss., 1882, pp. 529-531. Ueber 
den dritten Formbestandtheil des Blutes. 
3, 1874. Proceedings of the Royal Society, London, No. 153, pp. 
391-398 (with plate). An account of certain organisms occurring 
in the liquor sanguinis. 
4,1881. Sequin’s Arch, of Med., 1881. (Referred to in 2.) 
Pouchet.—1,1878. Gazette Medicate de Paris, 1878, p. 97. Note 
sur la reg6neration des hematies. 
2, 1878. Ditto, pp. 135-136. De l’origin des hematies. 
3, 1878. Ditto, p. 208. Note sur la circulation des rongeurs. 
4, 1878. Ditto, p. 316. Note sur 1’evolution des elements du sang 
des ovipares. 
5, 1879. Gaz. Med. 1879, pp. 184-185. De la degenerescence de 
la moelle des os. 
8. Gaz. Med. de Paris, 1879, p. 209. 
7. Revue scientifique, Sept., 1879, pp. 278-285. La formation 
du sang. 
Ranyier.—1. Traite technique d’PIistologie. Paris, 1876-1882. 
2, 1873. Gazette Medicale de Paris, 1873, pp. 93-94. Du mode 
de formation de la fibrine dans le sang extrait des vaisseaux. 
3, 1873. Comptes Rendus des Stances de la Societe de Biologie, 
1873, p. 46. 
Rauschenbach.—1, 1882. Tiber die Wechselwirkung zwischen 
Plasma und Protoplasma. Diss., Dorpat, 1882. 346 
GEO. T. KEMP. 
Riess.—1, 1872. Arch. f. Anat. u. Physiol., 1872, pp. 237-248 
(with plate). Zur pathologischen Anatomie des Blutes. 
2, 1879. Berliner klin. Wochenschrift, 1879, pp. 696-697. 
Bemerkungen iiber die Zerfallskorperchen des Blutes und ihr 
Verhaltniss zur Anamie. 
3, 1873. Centralbl. f. d. med. Wiss., 1873, pp. 530-533. Ueber 
sogenannten Micrococcen. 
Rindfleisoh.—1, 1878. Lehrbuch der pathologischen Gewebe- 
lehre, Leipzig, 1875. 
Schimmelbusch.—1, 1885. Fortschritte der Medicin, 1885, pp. 
97-103. Die Blutplattchen und die Blutgerinnung. 
2, 1885. Virchow’s Arch., Vol. 01, pp. 201-244 (with plate). 
Die Blutplattchen und die Blutgerinnung. 
Schmidt, A.—1, 1877. Die Lehre von der fermentativen Gerin- 
nungserscheinungen von der eiweissartigen thierischen Korperfliis- 
sigkeiten. Dorpat, 1877. 
2, 1875. Pfluger’s Arch. f. d. gesammt. Physiol., etc., Vol. XI, 
pp. 291-369. Ueber die Beziehung der Faserstoffgerinnung zu den 
korperlichen Elementen des Blutes. 
3, 1875. Ditto, pp. 515-577 (with plate). Continuation of 2. 
4, 1882. Archives de Physiologie, 2d series, Vol. IX, pp. 513- 
592. Recherches sur le role physiologique et pathologique des leuco- 
cytes du sang (especially Section 6). 
5, 1874. Pfiiiger’s Archiv, etc., Vol. IX, pp. 353-357. Ueber die 
Beziehung des Faserstoffes zu den farblosen und den rothen Blut- 
kbrperchen, und iiber die Entstehung der Letzteren. 
6, Dorpat Med. Zeitschrift, Vol. V, p. 257. Ueber die weissen 
Blutkorperchen. 
7, 1883. Arch, de Physiologie, 1883, pp. 112-122. Recherches 
sur les leucocytes du sang. 
Schultze, Max.—1, 1865. Archiv fur mikroskopische Anatomie, 
Vol. I, pp. 1-43. Ein heizbarer Objecttisch und seine Verwendung 
bei Untersuchungen des Blutes. 
Semner.—1, 1874. Ueber die Faserstofibildung im Amphibien- 
und Vogeiblut, und die Entstehung der rothen Blutkorperchen der 
Saugethiere. Diss., Dorpat, 1874 (67 pages and plate). 
Slevogt.—1, 1883. Ueber die im Blute der Saugethiere vorkom- 
menden Kornchenbildungen. Diss., Dorpat, 1883 (36 pages). 
Virchow.—1. Gesammelte Abhandlungen, 1856. 
Vulpiah.—1, 1873. Gaz. Med. de Paris, 1873, p. 94. (Abstract 
of communication to Soc. dg.BiolO “NEW ELEMENT” OF THE BLOOD. 
347 
2, 1873. Comptes Rendus des Stances de la Societe de Biologie, 
p. 49. 
Weigert.—1, 1877. Virchow’s Arch., Vol. LXX, pp. 461-490. 
Ueber Croup und Diphtheritis. 
2, 1880. Ditto, Vol. LXXIX, pp. 87-123. Ueber die patholo- 
gischen Gerinnungsvorgiinge. 
3, 1883. Fortschritte d. Medicin, 1883, pp. 406-414. Die neuesten 
Arbeiten iiber Blutgerinnung. 
4, Ditto, pp. 406-414. Continuation of 3. 
Wooldridge.—1, 1881. Arch. f. Anat. u. Physiol., 1881, pp. 
387-412. Zur Chemie der Blutkorperchen. 
2. 1881. Proceedings of the Royal Society of London, Vol. XXXII, 
pp. 413-418. The relation of the white blood corpuscles to coagula- 
tion. 
3. 1883. Arch. f. Anat. u. Physiol., 1883, pp. 389-393. Zur 
Gerinnung des Blutes. 
4. Journal of Physiology, Vol. IV, pp. 226-230. Further 
observations on the coagulation of the blood. 
5. 1884. Ditto, pp. 367-369. On the coagulation of the blood. 
6. Proceedings of the Royal Society, London, Vol. XXXVI, pp. 
417-420. On the origin of fibrin-ferment. 
7. Ditto, Vol. XXXVIII, pp. 69-72. A new constituent of the 
blood and its physiological import. 
8. Ditto, pp. 260-264. On the fibrin-yielding constituents of 
the blood-plasma. 
Zimmerman.—1, 1846-1848. Rust’s Magazin f. d. gesammte 
Heilkunde, Vol. LXVI. Ueber die Formgebilde des menschlichen 
Blutes, in ihrem nahern Verhaltniss zum Process der Entzundung 
und Eiterung. 
2, 1860. Virchow’s Arch., Vol. XVIII, pp. 221-242. Die Blut- 
korperchenfrage. 
DESCRIPTION OF PHOTOGRAPH. 
The photograph was taken from a specimen of dog’s blood pre- 
pared as follows : A cannula was introduced into the.femoral artery 
and the blood allowed to stream freely through. A cover-slip 
then held in a pair of forceps and touched to the stream of blood; 
without losing a moment a stream of .75 per cent. NaCI solution was 
directed obliquely against the slip, which was then instantly dropped 
into a watch-glass of 1 per cent. a*‘sen4o- acid. After about twenty 348 
GEO. T. KEMP. 
minutes it was taken out, washed and left all night in a 1 per cent, 
aqueous solution of Bismarck brown. It was then washed in water 
and mounted in a saturated solution of potassium acetate. 
The photograph is of especial interest as being the first ever obtained, 
and showing at once the following points: 
1. The form of the plaques (circular and oval). 
2. Their biconcavity (appearing as a light shade in the centre of 
the plaque). 
3. Their variation in size. 
4. Their tendency to adhere to each other and to the cover-slip 
within a few seconds. Their adherence to the cover-slip is shown 
by the red corpuscles in the field. These appear as circular discs 
with a bright centre and a halo around their periphery. This is due 
to their being out of focus, the objective being focussed on the 
plaques adhering to the slip, while the red corpuscles lie in the 
liquid between the slip and the slide. 
5. They are not fragments of disintegrated leucocytes, but are 
distinct histological elements. 
DESCRIPTION OF FIGURES, PLATE XIX. 
Figure 1.—Drawn from an unstained specimen prepared as 
described for photograph. The lenses used were Zeiss homo- 
geneous immersion obj., oc. 4 (x = 1450), with Abbe condenser, a, 
Plaques seen full on face, slightly or not at all altered, a', Plaques 
slightly altered in outline, showing the first changes toward breaking 
down, b, Plaque seen on edge, b', Plaque seen partly on edge 
and partly on surface, c, Group of three plaques, each seen in 
different position, d, Plaque fixed at one end and drawn out by 
currents. This is quite a common appearance. The large, round 
figure to the right is a red corpuscle drawn to scale for comparison 
in size with the plaques. 
Fig. 2.—Group of plaques enclosing three red corpuscles. The 
plaques have now become granular and of irregular outline. This 
is the stage in which the plaques are most commonly found when 
examined in .75 per cent. NaCl solution or Bizzozero’s fluid. Drawn 
with Zeiss homogen. ini., oc. 2 (x= 790), with x\bbe condenser. 
Fig. 3.—Same as Fig. 2, except that the plaques are more broken 
down, and farther on the way to form a typical granular mass. It 
will be observed that the plaques are not only jagged and granular, 
but they have begun to fuse together and lose their individual out- “NEW ELEMENT” OF THE BLOOD. 
349 
line. The group includes a red corpuscle and a tetranuclear leuco- 
cyte. Lenses same as in Fig. 2. 
Fig. 4.—Typical granular mass. The plaques are so far fused 
together that, for the most part, their individual identity is lost. 
Lenses same as Figs. 2 and 3. 
Fig. 5.—Typical granular mass with fibrin-threads radiating from 
it, and needle-shaped threads (crystalloids) deposited freely in the field. 
Fig. 6.—Drawn from specimen of scanty clot in diluted blood. 
Shows how fibrin is deposited freely in the field, as well as in con- 
nection with the granular masses. 
Fig. 7.—Thick formation of fibrin. Shows fibrin-threads branch- 
ing all over the field, enclosing granular masses, red corpuscles and 
a trinuclear leucocyte. Drawn from a specimen stained with methyl 
violet. 
Fig. 8.—Single plaques showing the “ bladder formation ” from' 
the action of water. The granules seen in the light part of the 
“ bladders ” are due to the method of preparing the plate. The dark 
part should be granular, as represented; the light part should be 
perfectly homogeneous. 
Figs. 9 and 10.—Plaques which had come together and afterwards 
swollen up from the action of iodine solution (aqueous). The 
peculiar shape was likely produced by pressure on the cover-glass. 
Fig. 11.—Same as Figs. 9 and 10. Drawn particularly for com- 
parison with Fig. 13, q. v. 
Fig. 12.—Leucocyte with adherent plaques which have swollen 
up and formed “ bladders.” From a specimen stained with methyl 
violet. 
Fig. 13.—After Lowit. Lowit gives the figure (especially c?) to 
illustrate the extrusion of plaques from the body of leucocytes, 
and I reproduce his drawing to compare with Figs. 11 and 12. (See 
text.) 
Fig. 14.—Leucocytes, a, b, c [and #(?)], multinuclear leucocytes 
supposed by Rauschenbach and Hlava to break down during coagu- 
lation, while the uninuclear variety cl do not break down. Notice 
the resemblance of the smaller nuclei in a, h, c and g to the plaques 
in Fig. 2. STUDIES FROM BIOL.LAB 
YOL.III.PLATE XIX. 
a hoe:n a. co. l,th.