THE PATHOLOGY OF UREMIC INTOXICATIONS. 
BY 
0. A. HERTER, M.D., 
Visiting Physician to the City Hospital, Consulting Physician to the Babies’ 
Hospital, New York. 
(Reprinted from Montreal Medical Journal, May, 1898.) THE PATHOLOGY OF URiEMIC INTOXICATIONS.1 
C. A. Herter, M.D., 
BY 
Visiting Physician to the City Hospital, 
Consulting Physician to the Babies’ Hospital, New York. 
In venturing to address you this evening on the subject of uraemic 
intoxications, I am not unmindful of the difficulties that are inseparable 
from the problem which I shall undertake to discuss. Indeed it 
cannot be denied that one of the attractions of this subject is the 
obscurity which surrounds the oft-recurring phenomena which we 
call uraemic—phenomena that have excited the interest and baffled 
the insight of Richard Bright, of Traube, of Frerichs, and of many 
lesser intellects. 
Although a study of uraemia that has extended over many years does 
not make it possible for me to isolate, as definitely as could be wished, 
the pathological factors concerned in the production of the symptoms 
of uraemia, I believe I can, at least, indicate the direction in which we 
shoulddook for an explanation of the nature of these symptoms. It 
will be my purpose to show that the clinical phenomena which we 
include under the term uraemia are dependent on toxaemic states rather 
than upon primarily mechanical causes. In doing this I shall make 
use of observations which I have accumulated from the study of the 
blood in 28 cases of uraemia, which have either been observed by me 
in ray medical service in the City Hospital, or have been placed at my 
disposal by interested colleagues. In addition to this, I shall draw 
upon facts which have come to my knowledge from a study of experi- 
mental uraemic states. It would be a satisfaction to present these 
facts in fnll detail, but as this could hardly be done without exhaust- 
ing your patience, a summary of results must suffice for the present 
1 Read before the Montreal Medico-Chirurgical Society, at McGill University 
March 4, 1898. 2 
purpose.1 For the sake of convenience, the facts to be surveyed may 
be grouped under the following subjects : the toxic properties of the 
blood in uraemic states as measured by intravenous infusion; the 
relation of uraemia to the urea of the blood, to the extractive sub- 
stances, and to the potassium salts ; cerebral oedema and uraemia ; 
the internal secretion of the kidney and uraemia ; experimental 
uraemia from double nephrectomy and its relation to human ob- 
structive uraemia, and other types of human uraemia. 
A question of fundamental importance in the pathology of uraemia is 
whether the blood in this condition is more toxic when introduced into 
the circulation of animals than is human blood from normal persons. In 
the hope of answering this question, the blood serum from 28 uraemic 
patients, who were bled during life, was introduced intravenously into 
rabbits, proper precautions being taken as regards the freshness of 
the serum, its temperature, the rate of infusion, the exclusion of air 
from the vein, etc. The observations were not confined to rabbits, 
but included in some instances dogs and monkeys. Two different 
methods of intravenous infusion were employed. In one case the 
serum was infused into the femoral vein at a fixed rate until the com- 
mencement of fatal symptoms. In the other case, the serum was 
injected in much smaller amount, either into the femoral vein or into 
an ear vein with a view to ascertaining the minimum dose capable of 
causing death in the course of 24 or 36 hours. A difficulty which at 
once confronts the investigator in the use of these methods is the fact 
that normal human serum is in itself toxic to rabbits in a considerable 
degree, owing chiefly perhaps to its power of inducing coagulation. 
Thus I have found that its requires from 25—40 c.c. of normal human 
serum to the kilo to initiate fatal symptoms in rabbits when the 
method of continuous infusion is employed, while it seems generally 
agreed among investigators that the fatal dose of human serum, as 
O O > 
employed by the second method, is from 9—12 cc. per kilo. Inas- 
much as it is found that there is considerable variation in the 
toxicity of the same serum for different rabbits of the same weight, 
notwithstanding every precaution in making the infusion, it is clear 
that it is necessary to observe much care in concluding that a given 
class of serums is more than normalty toxic.2 
Of the ursemic serums which were studied with respect to their 
2 Uhlenhuth claims that the serum should be subcutaneously injected into 
guinea pigs to avoid the error incidental to intravenous infusions. In several 
instances where I have employed this method, there seems to be no doubt that the 
blood was more than normally toxic, judging from the standard of normal toxicity 
given by Uhlenhuth. 
1 The details of the experimental observations will be elsewhere published. 3 
toxicity, 19 were obtained from cases of uraemia characterized by the 
occurrence of well marked and repeated convulsive seizures,1 six of 
these being instances of puerperal eclampsia. In the nine remaining 
cases the serum was obtained from patients with chronic nephritis, 
characterized by dyspnoea and high tension pulse—none of these 
patients having had convulsive seizures. The inferences which we 
may make from the study of these serums are as follows : The cases 
of nephritis characterized by dyspnoea and high tension pulse without 
convulsions, taken as a class, do not yield positive indications that 
the toxicity of the serum was greater than normal, although in at 
least two of the nine cases the toxicity was considerably greater than 
any I have observed in health. On the other hand the least toxic of 
these serums possess toxic values that come so near the highest toxic 
values observed in normal serums that we cannot be certain that they 
are quite normal. More observations are required before a final judg- 
ment can be formed regarding these cases. Taken as a class, the 
serums from the convulsive group of uraemias show a degree of 
toxicity distinctly greater than any which I have observed for normal 
serums, and although there are a few of the 19 cases in which the 
results are difficult to interpret, there seems to be no doubt that an 
increase in the toxic properties of the blood is a characteristic of cases 
of convulsive uraemia. This latter statement apparently holds good 
also of certain serums from patients with puerperal eclampsia, but 
here again more observations are required. Yolhardt has quite 
recently denied that there is any increase in the toxic properties of 
the blood of eclamptic patients, but his results cannot be regarded as 
final. 
Admitting that the serum in convulsive uraemia is more toxic than 
normal serum, the question at once arises, what is the cause of this 
pathological increase ? 
It is not possible at present to give a satisfactory answer to this 
question, but the following facts bear upon it. In the first place the 
toxicity of uraemic serum for animals is greatly reduced by exposure 
to moderate heat, say 60° C. for a few minutes. This is also true of 
the toxic properties of normal human serum. These facts suggest 
that the increased toxicity of uraemic blood is dependent on the 
presence of an increased amount of a toxic proteid substance which is 
normally present in the blood.2 We cannot, however, feel certain 
that the increased toxicity of the serum does not depend on some 
1 In one of these cases there were merely coarse twitchings never amounting to 
typical convulsions. 
2 The toxic properties of urea, extractives, salts, etc., are of course not influenced 
by so moderate an elevation of temperature. 4 
unknown foreign substance. A fact which strengthens the view that 
a proteid substance is responsible for the toxicity of uraemic, as of 
normal serum, is that the removal of the proteids by means of absolute 
alcohol almost wholly deprives the serum of toxic properties. 
In trying to form an estimate of the significance of an increase in 
the toxic properties of uraemic as contrasted with normal serums, it is 
important to recognize that such an increase in toxicity does not con- 
stitute proof that the cause of this augmented toxicity is the cause of 
the obtrusive cerebral symptoms of uraemia. It is conceivable that 
alterations may occur in the blood which possess little pathological 
significance for the human organism, but which are nevertheless 
capable of rendering the serum more toxic than normal when intro- 
duced into the circulation of rabbits This does not, however, seem a 
probable explanation of the phenomenon of augmented serum toxicity, 
and is especially at variance with the fact that our uraemic serums 
have, in some instances, exhibited very striking toxic properties when 
introduced into dogs and monkeys, which do not possess the peculiar 
susceptibility to normal human serum that is observed in the case of 
rabbits. It is much more rational to regard the increase in the toxic 
properties of the blood in certain, if not in all, cases of uraemia, as 
evidence of a toxaemia, which is in some way connected with the 
symptoms of the uraemic state. It is possible that this toxaemia is 
distinctive of uraemia and always of the same character, but there are 
certain differences in the behaviour of different serums which make it 
likely that the toxaemia is not always the same. It is, moreover, 
certain that the toxic properties of the blood are augmented in con- 
ditions not uraemic in nature, for example, in acute lobar pneumonia, 
in scarlet fever, etc. 
Urea being by far the largest and most important constituent of the 
urine and representing the terminal product of proteid metabolism, 
it is not singular that this substance should have been for a long time 
regarded as the main factor in the production of uraemia. It is thus 
of the highest importance to determine the actual relation of urea to 
uraemic states. The conclusions reached by me in reference to this 
relation are based upon a considerable number of determinations of 
urea in the blood of uraemic persons, of persons with other pathologi- 
cal conditions than uraemia, and of entirely normal persons. Many of 
these observations were made upon blood drawn during life, but these 
were supplemented by others upon post-mortem blood, after it was 
demonstrated both in man and in animals that the urea content of 
ante- and post-mortem blood corresponds closely in the case of blood 
taken from well preserved bodies. In addition to making seventy 5 
observations on the urea of human blood, an endeavour was made to 
determine to what extent chemically pure urea possesses toxic proper- 
ties. The results of these studies may be briefly stated as follows: 
(a) In a small proportion of typical cases of uraemia, characterized 
by the occurrence of convulsions, the urea content of the blood was 
found to be well within the normal limits, notwithstanding that in 
these same cases the blood serum was distinctly more toxic than 
normal. It is plain that in these cases there is no ground for looking 
with suspicion upon urea as a factor in producing symptoms. 
(h) In the majority of cases of uraemia, especially in those cases in 
which the secretion of urine has for a long time been scanty, the urea 
of the blood was distinctly increased, this increase sometimes reaching 
to sorlo or even 20 times the normal urea content. It is, however, 
a very striking fact that many cases of chronic nephritis were found 
to be characterized by a markedly excessive urea content, even where 
there were no symptoms that would ordinarily be called uraemic. This 
is merely a confirmation of an old observation by Bright and Christi- 
son which has not received the attention which it deserves. Among 
the cases in which the urea was increased in the absence of typical 
uraemic symptoms were 14 cases of acute lobar pneumonia with fatal 
termination. These cases make it clear that renal inadequacy is a 
feature of many fatal pneumonias. Whether such inadequacy is to 
be considered uraemic in the absence of the usual symptoms of uraemia 
will be considered later. 
(c) Experimental studies upon dogs and monkeys show that pure 
urea is toxic when infused intravenously, but only when very large 
quantities are employed, i.e., quantities equal to many times the daily 
urea excretion of the animal. The symptoms produced by such in- 
fusions are fairly constant in dogs, and consist in initial slower and 
stronger heart action, copious diuresis, diarrhoea, contraction of the 
pupils, irregular fibrillary contractions of the muscles, and finally 
general and severe tonic or clonic spasms and death. If we make the 
bold assumption that in human nephritis the susceptibility of the 
nervous system to the influence of urea is similar to that observed in 
dogs, we would expect convulsive seizures to arise in man when the 
urea content of the blood reaches about .5 per cent. It was found 
however, that in many of the cases of human nephritis, in which con- 
vulsions occurred, the percentage of urea was considerably below this 
point, while in a few cases in which this percentage was much ex- 
ceeded, convulsions were absent. 
It is, of course, obvious from the foregoing facts that urea does not 
play a necessary part in the causation of the symptoms of human 6 
convulsive uraemia. This does not, however, prove that the accumu- 
lation of the urea in the blood in large excess is of little importance. 
There is good evidence that such an accumulation points to a consider- 
able degree of degeneration of the secreting epithelium, and the 
degeneration which permits the storage of urea, permits the storage 
of other constituents of the blood, including sodium chloride, potassium 
salts, nitrogenous extractives, and possibly a toxic proteid substance. 
It is partly on account of this multiplicity of substances retained in 
the blood at the same time with urea that we are unable to fix more 
definitely on the part which urea plays as a toxic substance. It is on 
the whole very likely that urea in conjunction with the accumulation 
of other substances in the blood, as the result of renal insufficiency, 
exerts an influence in the production of ursemic symptoms—especially 
uraemic vomiting and diarrhoea. The urea of the blood has been 
greatly increased in all the instances of uraemic vomiting and 
diarrhoea that I have had an opportunity to study. 
It is desirable to indicate here that the customary determination of 
urea in the urine in cases of nephritis gives no more indication of the 
quantity of urea in the blood than does a mere record of the volume 
and specific gravity of the urine passed by such patients, and that it 
is hence an unnecessary procedure. If the kidneys are incompetent a 
large reduction in the urea excreted is indicated sufficiently by the 
reduction in the total solids, this being expressed by the volume and 
specific gravity. If there is only a moderate daily reduction in the 
urea excretion this cannot be discovered by the ordinary methods 
used. To determine this it would be necessary to keep a daily record 
of the nitrogen taken into the body, of the nitrogen of the urine and 
of the fseces, and also of the body weight. It is easy to see how, in the 
course of a few months of mild renal incompetency, the urea of the 
blood would increase enormously if the kidney lagged only to the ex- 
tent of a few grams of urea daily—too small an amount to miss by 
ordinary examinations. It is no exaggeration to state that for prac- 
tical purposes the consideration of the volume and specific gravity of 
the urine yields all the information that can be obtained by determin- 
ing the urea of the urine. What we really require is an estimate of 
the urea of the blood in our nephritic patients, and this is entirely 
practicable, for only a few cubic centimetres of blood are necessary for 
this purpose. I have no doubt that this procedure will in time be 
frequently employed, as the urea content of the blood is a most 
important indication of the competency or incompetency of the 
kidneys. 
In recent years there has been some disposition to regard the 7 
extractive substances as the cause of uraemic symptoms, but the evi- 
dence in support of this view has never been strong. The results of 
a personal study of this question may be briefly referred to. The ex- 
tractives of the blood, that is, the substances which can be extracted 
by means of ether and alcohol, wrere determined in more than one 
hundred instances, including normal human and dog’s blood, blood 
from nephritis with and without uraemic symptoms, from septicaemia, 
acute lobar pneumonia, etc. It was found that there is no definite 
relation between uraemia and the quantity of extractives in the blood. 
In general the extractives are somewhat increased where the quan- 
tity of urea is increased, but there are cases of uraemia where the ex- 
tractives are apparently normal in amount, and there are cases which 
would not usually be classed as uraemic where the extractives arc 
markedly increased. The evidence indicates that while extractive 
substances in excess cannot be regarded as entirely harmless for the 
organism, they certainly cannot be looked upon as playing other than 
an auxiliary part in the production of uraemic symptoms. In this 
connection it may be stated that Dr. A. J. Wakeman, at my request, 
made a series of laborious observations on the blood of uraemic 
patients with the use of the Otto-Stas method, for the purpose of 
isolating any alkaloidal substances which might exist there. The in- 
jection into guinea pigs of the material recovered by the Otto-Stas 
method yielded wholly negative results. 
The conclusions that have been stated regarding the extractives of 
the blood are applicable to the potassium salts theory of uraemia, which 
was originally advanced by Felz and Ritter, and which constitutes a 
most instructive chapter in the history of theories of uraemia. IN uraer- 
ous observations made by me confirm the statement of Horbaczewski 
that the content of potassium salts in uraemic blood may be quite normal 
in amount. This seems to be especially true of the blood of puerperal 
eclampsia. These salts are, however, distinctly increased in many in- 
stances of uraemia, but apparently never enough to make them wholly 
responsible for grave nervous symptoms. The potassium theory as 
an exclusive cause of uraemia has recently been revived in France by 
Charrier, but upon wholly insufficient grounds. It may be said at 
present that, while the potassium salts cannot bo considered to play a 
leading part in the production of uraemic symptoms, their presence in 
excess in the blood must be regarded as a possible factor in precipitat- 
ing symptoms of intoxication. 
The ammonium carbonate theory of Frerichs, once so popular, has 
now only a historical interest and need not be discussed here. An 
allied hypothesis has, however, been suggested, namely, that the cause 8 
of uraemia is tlie presence in the blood of the ammonium salt of car- 
barn ic acid—ammonium carbamate. The recent investigations of Hahn 
Pawlow, Massen and Nencki render it probable that urea is formed in 
the liver by the dehydration of ammonium carbonate, and this fact 
has led some physiologists to suspect that the highly toxic ammonium 
carbamate may be responsible for uraemic states. This view appears to 
me to be at variance with the following facts: 
1. In watery solution ammonium carbamate is very unstable and is 
rapidly converted into ammonium We know, however, 
that ammonium carbonate does not occur in uraemic blood in sufficient 
amount to produce symptoms, and usually cannot be found at all. 
2. The toxicity of uraemic blood is not lessened by dialysis, as it 
would be if the toxicity depended on a diffusible ammonium salt; nor 
does the diffusate contain ammonium. 
3. The urine of uraemic patients does not necessarily contain an in- 
creased proportion of N. of ammonia, as it should do if the synthesis 
of urea were impaired. On the other hand, in liver diseases in which 
there is extensive parenchymatous destruction, the N. of ammonia 
may be greatly increased, even in, the absence of uraemic symptoms. 
Owing to the instability of ammonium carbamate its isolation from 
the blood is impracticable, and inferences as to its occurrence there 
depend chiefly on indirect] evidence, The conclusion seems justified 
that such knowledge as we possess does not support the supposition 
that ammonium carbamate is concerned with the production of uraemic 
intoxications. 
Before passing to the more constructive consideration of the uraemic 
problem, it is proper to make brief reference to two widely different 
theories of the nature of uraemia. One of these is the celebrated 
hypothesis of Traube that renal disease causes thinning of the blood 
plasma, hypertrophy of the left ventricle and excessive arterial pres- 
sure. If the arterial tension is increased beyond a certain point or 
the plasma of the blood becomes further thinned, oedema and anaemia 
of the brain are produced and uraemic symptoms result. There are 
fatal objections to this theory. These are : 1. That there may be 
marked cerebral symptoms without arterial tension. 2. That the 
specific gravity of the serum is often normal in typical uraemia. 
8. That there are uraemic patients in whom neither cerebral anaemia 
nor cerebral oedema are found at autopsy. 4. That a marked degree 
of anaemia of the brain and of oedema is occasionally found in the ab- 
sence of all symptoms resembling uraemia. 
It is, however, clear from clinical study that there is often a close 
association between certain uraemic symptoms, especially convulsions 9 
and dyspnoea and high arterial tension. Nor is there any doubt that 
when we reduce excessive vascular tension by means of vaso-dilators, 
such as nitro-glyceririe, we often relieve the symptoms in the case of 
dyspnoea, sometimes in a striking degree. At present it appears prob- 
able that the high arterial tension of uraemia is due to the action of 
toxic material in the blood. It is not at all inconsistent with this view 
that variations in the circulation of the central nervous system should 
influence, in important ways, symptoms which, like dyspnoea, are of 
central origin. 
Although we cannot accept the theory of Traube as advanced by 
him, the appearances observed in the central nervous system in persons 
dying of uraemia1 (especially the loss of chlorophyllic substance seen 
in nerve cells) indicate that mechanical conditions, such as local 
anaemia, congestion or oedema, have been operative in damaging the 
nerve elements. It seems probable that these conditions are not 
primary, but depend on toxaemic states. 
Another theory of uraemia which cannot be passed by, is that of 
Brown-Sequard, who holds that the kidney elaborates an internal 
secretion which is essential to health, and the suppression of which is 
largely responsible for the phenomena of uraemia ; the accumulation 
in the blood of toxic substances which should be excreted by the urine 
having little or no influence on the causation of these phenomena. 
This conception of uraemia rests mainly on the alleged fact that the 
injection of kidney extract into the circulation of a nephrectornized 
dog causes the temporary disappearance of uraemic phenomena. The 
evidence which has been advanced to establish this fact must be 
regarded as insufficient. Thus the observation of Brown-Sequard 
that the injection of kidney extract causes an increase in the muscular 
power of nephrectornized dogs, rests on a small number of experiments 
which appear neither to have yielded decided results nor to have been 
subjected to careful control, and Meyer’s contention that the injection 
of renal extract gives marked relief to the dyspnoea of double nephrec- 
tomy likewise rests on a small number of observations that seem 
distinctly to call for proper controls. Nor can we place much reliance 
on the claims of Teissier and Frenkel, in a recent publication, that 
the injection of a few cubic centimeters of renal extract in the human 
uraemic subject is capable of rendering a hypotoxic urine hypertoxic 
by stimulating the elimination of toxines through the urine at the 
same time that the symptoms of uraemia are ameliorated. Consider- 
1 These changes are not by any means distinctive of uraemia, but occur in a variety 
of conditions. According to Dr. James Ewing they are best seen where the nerve 
elements have been subjected to pressure, as in cerebral hemorrhage. 10 
able personal experience makes me highly skeptical as to the propriety 
of our (h awing inferences as to the condition of the blood from the 
effects of intravenous infusions of urine in animals. A very obvious 
and serious gap in the experiments of Teissier and Frenkel is the 
absence of observations on the toxic salts ingested with the food and 
eliminated with the urine, the toxicity of the urine, both in health 
and disease, being largely dependent on its potassium salts. Future 
investigations may show that the kidney elaborates an internal 
secretion, but at present we are justified in taking the position that 
the observations now relating to this question do not help us in ex- 
plaining the pathology of uraemia. 
Passing now to a consideration of the clinical types of uraemia with 
a view to the discussion of their pathology, it is desirable first to 
make reference to the phenomena of double nephrectomy in dogs and 
their relation to human uraemia of obstructive origin. 
The alterations in the composition of the blood that are entailed by 
double nephrectomy are of the greatest interest in the study of the 
pathology of uraemia, for they necessarily represent the results of the 
most extreme degree of renal incompetency per se and without com- 
plicating factors such as are commonly present in human uraemia. 
The following description of the symptoms and pathological altera- 
tions incidental to experimental uraemia, is based on a series of 10 
successful cases in dogs, as well as on cases in the pig and rabbit— 
and also upon a number of instances in which the ureters were tied 
upon both sides. It may be stated at the outset that the symptoms 
were essentially the same in the case of double nephrectomy as in 
ligation of the ureters. These symptoms consist, in a typical case, of 
moderate prostration following the operation, of repeated vomiting1 
sometimes associated with diarrhoea, of slow and deep respiration, of 
slow, full and high tension pulse and, not rarely, of fibrillary twitch - 
ings. In only one instance did true convulsive seizures occur. Death 
is usually preceded by a period of drowsiness or actual coma. In 
none of my animals was the operation survived more than four and a 
half days, and most of them lived less than three days. In cases 
unaccompanied by infection the temperature is generally one or two 
degrees below normal for one or two days before death. 
A considerable number of observations have been made by me to 
determine whether the blood of nephrectomized dogs is more toxic to 
rabbits than the blood of normal dogs ; or, to put it a little differently, 
1 The vomiting referable to the removal of the kidneys must be distinguishable 
from that which results from peritonitis accompanying accidental infection in these 
cases. 11 
to determine whether the experimental uraemia of double nephrectomy 
is comparable with human uraemia of the convulsive type as regards 
the toxicity of the blood. There can be little doubt that this is a 
most important question in the pathology of uraemia, and I regret 
that it is not possible for me to make an unqualified statement in 
regard to it. Owing to the crudeness of our methods of studying the 
toxic properties of the blood, it is not possible to detect moderate 
deviations from the normal toxicity. It can, therefore, only be said 
that there does not seem to be a marked difference in the toxic 
properties of the blood within 48 hours after nephrectomy as com- 
pared with the blood of the same animal previous to nephrectomy, 
but that the toxicity of the blood seems to be increased in dogs that 
live a longer period. 
As regards the changes in the chemical composition of the blood, 
our information is much more positive. We know, for example, that 
the urea of the blood is remarkably increased after double nephrectomy 
—often reaching ten times the normal percentage at the end of 
three days. The extractives are also distinctly increased. A moder- 
ate increase in the total salts of the blood is probably a regular feature 
of the blood of nephrectomized animals. The potassium salts may be 
somewhat increased, but on the other hand may not be appreciably 
changed. The total proteids undergo no alteration in amount. A 
very interesting feature of the blood has been the marked increase in 
fibrin which was noted in a number of the nephrectomized dogs. This 
observation, though one of much interest if confirmed, has not yet 
been subjected to the controls which are necessary to establish it as a 
fact, for a very definite source of error remains to be eliminated. This 
consists in the fact that a part of the increase in fibrin noted after 
nephrectomy may be due to the bleeding which was practised several 
days before nephrectomy for the purpose of establishing a basis of 
comparison before and after operation, for it is known that the fibrin 
of the blood is increased by bleeding.1 
In order to be able to compare the symptoms of double nephrec- 
tomy with the obstructive type of human uraemia, the cases of 
1 Since this paper was read I have succeeded in several instances in performing 
double nephrectomies and ureter ligations without losing more than a few c.c. of 
blood, through hemorrhage. All these animals have shown a doubling of the normal 
fibrin content of the blood after a few days. A dog subjected to laparotomy without 
nephrectomy, but with a moderate loss of blood, showed no increase in the fibrin 
content of the blood. 
These observations seem of especial interest in connection with the altered 
toxicity of the blood after nephrectomy, but other sources of error remain to be 
eliminated before the increase in fibrin can be positively attributed solely to the sup- 
pression of renal functions. 12 
prolonged anuria from obstruction accessible in literature were 
subjected to analysis. Of the 41 apparently reliable cases in which 
anuria lasted four days, 35 occurred in males. This striking differ- 
ence in sexual incidence is probably explained by the nature of the 
obstruction to the escape of urine. In 14 of the 21 cases in which an 
autopsy was made the ureter or pelvis of the kidney was obstructed 
by a calculus, on one or both sides, the obstruction thus created being 
the cause of the anuria. I have been unable to find the record of a 
case in which anuria in a female was due to obstruction of the ureter 
or pelvis by a calculus. Of 36 cases in winch there was either absolute 
anuria (if we can trust the histories) or in which only an insignificant 
quantity of urine was passed, the condition in 11 cases lasted more 
than 4 days and less than 7, in 18 cases lasted from 7 to 14 days, and 
in 7 cases lasted longer than 14 days. 
Although the records are not wholly satisfactory, they serve to 
bring out clearly some important facts regarding the symptoms of ob- 
structive uraemia. It is interesting to note that in seven of our forty- 
one cases it is definitely stated that no uraemic symptoms occurred, 
although some of the cases were of considerable duration (5, 5, 6, 7, 8, 
9 and 11 days). Although it is not unlikely that some of the unob- 
trusive indications of uraemia were overlooked in these cases, it is fair 
to suppose that there were no well-marked uraemic symptoms. It 
seems clear that it is the rule for uraemic symptoms not to begin for 
several days after the commencement of the anuria. In a number of 
cases more than a week elapsed before pronounced indications of 
uraemia began. In twelve of the forty-one cases it is noted that 
vomiting was present at some period of the anuria. In one case (that 
of Russell, lasting twenty days) it is said that vomiting was present 
from the beginning. Diarrhoea does not appear to be so frequent a 
symptom as vomiting. It was noted in only six cases. Headache was 
described in only six cases. Insomnia, with restlessness, was observed 
in several instances, and may be present from the beginning. Mus- 
cular paralysis was not recorded in any of the cases, although a con- 
siderable degree of muscular prostration was repeatedly observed, and 
is probably a relatively early symptom in most instances. Pronounced 
delirium was a rare symptom. General convulsions is another uncom- 
mon symptom, having been noted in only five of the forty-one cases. 
Twitchings of the muscles, not sufficiently wide in range to constitute 
convulsions, were observed in eleven cases. According to Roberts 
this is a highly characteristic symptom of obstructive anuria. 
As regards the state of the mental faculties it seems safe to say 
that death is usually preceded by drowsiness, if life lasts more than 13 
a week, but that the patient may in most instances be roused at any 
time. 
In four cases a urinous odor of the breath was noticeable and in one 
of these the skin also had a urinous odor. In one instance the breath 
is described as having an amrnoniacal character. In one patient, not 
included in tire collection, the suppression lasted four days and the 
skin of the neek and face was covered by crystals of urea. An im- 
portant clinical feature of obstructive uraemia is that the temperature 
is seldom elevated. In only one case is there a record of any fever, 
and in this the rise was slight. It is clearly the rule for the tem- 
perature to remain normal throughout or to be a little subnormal dur- 
ing the last days of life. 
On comparing the symptoms of these two sets of cases, the human 
obstructive uraemia and the experimental uraemia following double 
nephrectomy or bilateral ligation of the ureters, several important 
resemblances become apparent. Thus vomiting is an early and fre- 
quent symptom, while diarrhoea, though not rare, is distinctively less 
common. In both groups of cases marked muscular prostration is 
usual from the beginning. Occasionally, however, there is early rest- 
lessness. Indications of delirium are absent in the experimental as in 
the human cases and paralyses have not been observed. In the ter- 
minal stage, fibrillary tremors are common in both the human and the 
canine cases, while general convulsions are exceptional. Terminal 
coma may occur in either group, but consciousness can usually be 
aroused at any time. An important clinical resemblance lies in the 
fact that the temperature is either normal throughout or slightly sub- 
normal. There are, however, some points of difference. Thus, a 
patient with both ureters obstructed may live more than two weeks, 
while a dog with both ureters tied or with both kidneys extirpated 
lives less than one week. We can hardly attribute this difference in 
the duration of life to the shock of operation. It may depend on the 
activity of the skin in man. The amrnoniacal breath of human 
patients depends doubtless on the decomposition of urea in the gastro- 
enteric tract and the odor of the skin arises from the decomposition of 
urea in the sweat. It may happen that a greater accumulation of urea 
occurs in the blood in man than in the dog, owing to his longer sur- 
vival, and that this occasions the excretion of urea by the gut in the 
case of man. Notwithstanding these clinical differences, it seems 
probable that the pathological conditions which are responsible for the 
symptoms in nephrectomized dogs are essentially those that are 
responsible for the symptoms of obstructive uraemia—namely, the 
accumulation in the blood of urea, extractives, inorganic salts, and 14 
perhaps of a toxic proteid material the nature of which is at present 
unknown. 
Coming now to the consideration of other types of human uraemia, 
we find ourselves upon uncertain ground when we try to bring tire 
various clinical phenomena into relation with the pathological state or 
states which constitute their basis. This is because the pathological 
knowledge which we possess is still exceedingly meagre and probably 
inadequate to form the basis of a permanent pathological classification 
of the different combinations of systems which we call uraemic. There 
are, however, certain facts, some of which have already been alluded 
to, that seem to me to help us in the interpretation of the phenomena 
of human uraemia, and to these I wish to direct your attention. 
There is a well recognized group of patients with chronic nephritis 
whose leading characteristics clinically are high arterial tension, 
dyspnoea and slight albuminuria, and sometimes headache. Although 
the kidneys of such patients present considerable variations in their 
gross character, there is in all cases widespread degeneration of the 
secreting tubules, fibrous changes in the tufts, and a distinct increase 
in the intertubular connective tissue. Cases of this character are 
often benefited by venesection (at least temporarily) and it has thus 
become possible to obtain the blood for study in a number of instances. 
As already stated, it is not possible to say whether or not the serum 
from such patients is regularly more toxic than normal, although it 
appears as if this were the case in some instances at least. It is usual 
in these cases for the blood to contain an increased percentage of urea, 
thus affording a positive indication that the kidney is not wholly com- 
petent to perform its excretory functions. Cases of nephritis of this 
type may be the distant consequences of infection, but there is no 
reasorj to think that pathogenic bacterial products are present in the 
blood at the period when these cases run a chronic and entirely afeb- 
rile course. The only obvious pathological condition of the blood that 
is likely to be connected with the characteristic exacerbations of 
dyspnoea and the increase in arterial tension, is the retention in the 
blood of constituents that should be excreted, perhaps including an un- 
known toxic proteid material. In other words (leaving aside the fact 
that the water of the plasma may be increased in such cases) the con- 
dition of the blood in the cases described is probably analagous to the 
condition which was described as characteristic of double nephrectomy, 
and which, in all likelihood, is the basis of human obstructive uraemia. 
At least this much is certain—in the three different conditions which 
we have considered, double nephrectomy, obstructive uraemia and 
chronic nephritis, with high tension and uraemic dyspnoea—there is an 15 
actual retention of urea in the blood and not improbably an increased 
toxicity of the blood due to a proteid constituent. It is, of course, ob- 
vious that there are clinical differences between human obstructive 
uraemia and the dyspnoeic type which this pathological conception 
does not explain. We must, however, remember that in the one case 
we have to do with an acute condition arising sometimes in persons 
whose vascular system is not markedly altered, while in the second 
case the toxaemia is a chronic state arising in a person who, simul- 
taneously, and for reasons not understood, has developed cardio- 
vascular fibrosis. It is possible that this cardio-vascular fibrosis plays 
a mechanical part in the production of dyspnoea in the presence of a 
uraemic toxaemia like that referred to. 
In the course of the arterio-sclerotic type of chronic nephritis, with 
uraemic dyspnoea, etc., as well as in other types of nephritis, gastro- 
enteric disturbances, especially nausea, vomiting, and diarrhoea of an 
intractable character, are occasionally observed. It will be remembered 
that vomiting and diarrhoea are features of experimental uraemia from 
nephrectomy and of obstructive uraemia. 
There can be no doubt that these symptoms are due to a retention 
uraemia in which urea, extractives, etc., accumulate in the blood in 
large excess and are finally excreted by the gastro-enteric mucous 
membrane, causing diarrhoea, and in which vomiting is caused by the 
action of these toxic substances upon the medulla. I cannot under- 
take to say whether the known retained constituents of the blood 
(urea, extractives, salts) are in themselves responsible for these symp- 
toms or whether unknown substances contribute to determine these 
symptoms. There are several facts which are of significance in this 
connection. One is that the injection of urea into the blood eventu- 
ally causes diarrhoeal discharges containing urea. Another fact is that 
in every case of uraemia that I have studied in which uraemic vomit- 
ing or diarrhoea has been a feature, the blood has contained a marked 
excess of urea, etc. Again we find that vomiting and diarrhoea are 
characteristic symptoms both of obstructive uraemia and of double 
nephrectomy. We know that in the former cases crystals of urea 
may be found on the skin and that in the latter urea may be found in 
the intestine. Considering these facts together we cannot but feel 
justified in believing that the known retained constituents of the blood 
play a part in the production of the digestive symptoms of uraemia, 
possibly a leading part. 
In the course of a small proportion of cases of nephritis, convulsions 
constitute an obtrusive occurrence. The convulsive seizures may 
form part of the history of almost any type of nephritis—of nephritis 16 
with preponderant parenchymatous change to nephritis with extreme 
connective tissue alterations. In a certain number of cases the 
symptoms previous to the convulsive seizures have been those of the 
type already referred to, with high tension pulse, dyspnoea, moderate 
albuminuria, etc. In other words, this type of uraemia is liable at any 
time to become modified by spastic phenomena. It should be noted 
that in some of these cases, where convulsions are thus superposed, 
the temperature remains normal until the seizure, and is then only 
slightly elevated—not more than we might expect from violent 
muscular action. How do these cases differ pathologically from the 
high tension type of uraemia without convulsive seizures ? At the 
beginning of this paper reference was made to the fact that the 
toxicity of the blood was found to be markedly increased in many of 
eighteen cases of convulsive uraemia and at least apparently increased 
in all. It was found that the toxic properties of the serum in this 
group of cases were more pronounced than in the non-convulsive 
group with high tension. Moreover, it was found that in some of 
these convulsive cases the urea was not increased beyond the normal 
percentage. In a very few instances the urea, the extractives and the 
potassium salts were apparently within normal limits. It is plain that 
if the convulsions in these cases are of toxaemic origin they must 
depend upon some other substance than urea, and it is not unlikely 
that they depend on the presence of the proteid substance to which 
the exaggerated toxicity of the blood appears to be due. There does 
not, however, appear to be any significant difference between the 
convulsive and the non-convulsive group of cases, the difference in 
the toxicity of the blood being probably one of degree rather than of 
kind. Again, the clinical facts lend support to the view that the 
pathological basis of the two sets of cases cannot be very different, for 
the spastic phenomena are often so slight as to constitute only fibrillary 
twitchings that recur infrequently and bring about no noticeable 
change in the condition of th.e patient. It may be that the presence 
or absence of these nervous phenomena is connected with slight 
variations in the degree of toxaemia or with temporary alterations in 
the circulation of the brain. 
The evidence thus seems to favour the view that in the group of 
cases which has been discussed, the symptoms are dependent largely 
upon renal insufficiency and upon alterations in the blood that are 
secondary to this condition. It is important to realize that although 
a kidney may be competent to excrete urea so actively as to prevent 
the accumulation of urea in the blood, it does not necessarily follow 
that it is competent to excrete, or to transform and excrete other sub- 17 
stances which a healthy kidney would not permit to remain in the 
blood.1 
It may be that further studies will show that the essential patho- 
logical element in the forms of uraemia already considered is the 
presence of the proteid serum poison to which reference has been 
made; although in the form of uraemia characterized by gastro-enteric 
derangements, the accumulation of urea, salts, etc., appears to be a 
regular and probably a determining factor. 
An element quite different from simple renal insufficiency enters 
into many cases of uraemia, namely that of infection. 
There is a small but suggestive group of patients who, after ex- 
posure to cold or wet, develop high fever, partial suppression of urine, 
albuminuria and perhaps haematuria, with headache, delirium and 
coma. The peculiarity of these cases is that the kidneys have pre- 
viously been normal so far as can be determined by clinical methods. 
There seems little doubt that the cerebral symptoms in such cases 
of acute degeneration of the kidney or acute exudative nephritis are 
due wholly to the action of toxines and not to the retention of sub- 
stances in the blood which are normally eliminated by the kidney. 
These unusual but instructive cases appear to me to represent the 
type of uraemia most widely removed from human obstructive uraemia 
in its pathological basis. As might be expected, these cases retain 
their purely infective type only a short time, for the damage to the 
kidney soon leads to pronounced insufficiency and to the accumulation 
of urea, extractives, etc., in the blood in marked excess. A condition 
analogous to that just described can be produced in monkeys by the 
subcutaneous injection of pathogenic bacterial filtrates. 
In a considerable number of cases of chronic nephritis which have 
run an afebrile course there is a sudden development of fever, partial 
suppression with increase in the albumen of the urine and cerebral 
symptoms such as delirium, coma or convulsions. At autopsy the 
kidney may show the lesions of an acute nephritis grafted upon those 
of a chronic nephritis. Post-mortem cultures made from the blood 
and various organs frequently show the presence of pyogenic or other 
pathogenic bacteria. In short we have in these cases both clinical 
and pathological evidence of the occurrence of an acute infection. It 
seems reasonable to suppose that many of the symptoms which we 
call uraemic in these terminal cases are due to the combination of this 
infection with a pre-existing toxaemia due to chronic renal insuffi- 
1 The increase in the fibrin content of the blood which I have found in some 
cases of this sort is of interest in this connection, though its significance is still un- 
certain . 18 
cioncy. But such infections are by no means always terminal states. 
It has been shown by Welch and others that patients with chronic 
nephritis are especially susceptible to infection, and it often happens 
that a patient with chronic diffuse nephritis develops grave cerebral 
symptoms at the time of a trivial infection which causes a tonsilitis, a 
slight bronchitis or an otitis,—symptoms which we very properly 
look upon as uraemic, but from which there is apparently complete 
recover}^. 
Reference has already been made to the fact that the urea content 
of the blood was found to be almost regularly increased in persons 
dying of acute lobar pneumonia, in other words, that renal insuffi- 
ciency for urea is a characteristic of fatal pneumonias. This observa- 
tion suggests the question whether we are to regard a markedly ex- 
cessive accumulation of urea in the blood as an indication of uraemia 
even in the absence of typical clinical indications of uraemia. I 
strongly incline to the view that we should extend our conception of 
the term uraemia to include every case of renal insufficiency for urea 
although well defined uraemic symptoms be wanting. It has been 
made clear that typical uraemic symptoms may arise in persons whose 
blood shows no increase in urea, but this fact does not deprive the 
accumulation of urea, salts, etc., of clinical significance ; it merely 
illustrates that the pathological basis of what is clinically termed 
uraemia is not always the same. It seems to me desirable that we 
should regard any toxaemia as uraemic which can be shown to depend 
on the incapacity of the kidney to perform the functions of a 
healthy kidney, whether these functions consist simply in the elimina- 
tion of substances as they exist in the blood furnished by the renal 
artery, or whether they shall be shown also to consist in the trans- 
formation of certain elements of the blood previous to elimination. 
Again it is only rational that we should recognize that the essential 
elements of a uraemic intoxication may exist without being present in 
such a degree as to cause obtrusive and typical uraemic symptoms. 
Or, to restate the fact in a different form, we should recognize that 
there is such a thing as a latent urcemic intoxication. Such a latent 
uraemia is probably present in many forms of disease, especially acute 
disease, such as pneumonia, where the kidney is the seat of lesions, 
aud in chronic nephritis. In the former condition it constitutes a 
complicating state. 
The fact that such a toxaemia may be masked by associated condi- 
tions or may be in itself unrecognizable clinically does not prove that 
it is a state which exerts no influence in determining prognosis. 
In conclusion a word must be said about the most obscure type of 19 
uraemia—puerperal eclampsia. Efforts have been made to connect this 
state with the formation of toxic products formed by the chemical 
activities of the living cells of the embryo, with the absorption of 
toxic material formed in the intestine and with the accumulation of 
urea in the blood as the result of nephritis or of pressure on the renal 
vessels, but the efficacy of these supposed agencies still remains 
unproved. 
My personal experience with puerperal eclampsia is limited to the 
study of the blood of six victims of this state.1 In at least three of 
these cases the urea of the blood was not increased in percentage. It 
seems highly probable that the toxicity of the blood was distinctly in- 
creased in at least two of these cases. Of the other cases it cannot be 
positively stated that the blood was more toxic to animals than is ever 
the case with the blood of non-eclamptic puerperal women, nor, on the 
other hand, can it be stated that the blood was not more toxic than 
normal. 
At the present time there is a controversy as to the toxicity of the 
blood of eclamptic women, which can be definitely settled only by 
numerous and very carefully conducted observations. Although there 
is thus considerable uncertainty as to whether an increased toxicity of 
the blood is an essential feature of puerperal eclampsia there is im- 
portant indirect evidence of the existence of such a toxaemia. This 
consists in the presence of anaemic and hemorrhagic areas of necrosis 
in the livers of women dying of eclampsia.2 
Schmorl, who first described these striking lesions, regards the 
thromboses of the capillaries and small periportal veins with which 
they are associated as dependent on the passage into the blood of 
placental elements and products of placental degeneration. 
Flexner has succeeded in producing similar alterations in the liver 
by means of experimental intoxications, and there can be little doubt 
that we must regard the necrotic changes in the organs of eclamptic 
women as dependent on a toxaemia. How this toxaemia arises and 
how it is related to the toxaemias of nephritis already discussed 
remains to be discovered. 
Although this sketch of the pathology of uraemic conditions, made 
from a somewhat personal standpoint, shows us to be in possession of a 
meagre fund of knowledge respecting the pathological basis of uraemia, 
we may confidently hope for further enlightenment from experimental 
pathology. It seems to me that future researches should have refer- 
1 The blood from these patients was obtained through the courtesy of the attend- 
ing physicians of the Lying-in Hospital of New York. 
2 I have never met with these lesions in the livers of persons dying from other 
forms of uraemia than puerperal eclampsia. ence especially to the following topics; The toxic properties of the 
blood of uraemic patients, the physiological and chemical changes 
induced in the blood by nephrectomy, and the influence of intoxications 
of intestinal origin upon the normal organism and upon organisms 
which are the seat of nephritis. That the state of bacterial activity 
in the intestine is capable of exerting an important influence upon 
uraemic conditions is suggested by the observation which I have made 
that the albuminuria of a dog with chronic nephritis can be strikingly 
increased by feeding with cultures from the stools of entero-colitis 
It is also suggested by the exacerbation of symptoms which we some- 
times observe clinically in human patients after gross errors in diet.1 
This relation deserves further attention, as it is of the utmost practical 
importance in chronic uraemias. 
I had hoped to refer this evening to methods of treatment in 
uraemia, but am conscious of having already overstepped the limit of 
time imposed by reasonable usage. 
1 Nephritis and uraemia may also arise in children as a consequence of intense 
entero-colitis,