pitals made by Kapsammer, 750 cases of pyelitis and pyelonephritis were found, that is 3.6 per cent. And less than one-third of these were correctly diagnosed during the life of the patient. This is proof beyond question of the frequency with which these cases occurred and were unrecognized, and I believe it is no exaggeration to assert that nearly as large a proportion still go undiagnosed.

I shall first consider the recognition of this condition. And for our purpose we may omit that portion of the cases now usually recognized, namely, those in which there are definite localizing symptoms, such as pain and tenderness in the region of the kidney. These are so obvious that they cannot well be overlooked. Such cases, whether complicated by calculi or not, I believe to be decidedly the exception, not the rule. More often the disease appears in one of two quite different forms. In the first of these there is evidence of an acute infection, chills, a high fever which may be septic in type, rapid pulse, and a moderate leucocytosis. The localizing symptoms are lacking or so slight as to be overlooked. For some reason, perhaps by chance, I have seen more of these cases during the puerperium when they are particularly misleading, being naturally mistaken for an infection of the uterus. And I have been surprised to see in these not only the lack of pain in the kidney region but also the very slight tenderness here upon pressure which they show. It may be incorrect to speak of these as acute. Many of them are probably an acute exacerbation of a previously existing unrecognized pyelitis, with a sudden absorption of toxins from the inflamed area. However that may be, the condition is readily overlooked until the urine is examined. And I have found that one cannot always rely upon the statement of the attending physician that the urine shows nothing. I feel safer to examine the sediment myself.

The other type of case is the opposite of this. Instead of the marked constitutional symptoms the process here is so insidious as to be overlooked. The localizing symptoms, if present, are of that indefinite variety which may accompany any one of several different pathological conditions. This is the variety

which, if noticed at all, is liable to go into the category of lumbago, chronic appendicitis, etc. Here, as in the former type, it is the examination of the urine which must be the real determining factor in the case. And it is the examination of this with its positive findings which leads to the recognition of a very slight tenderness in the region of one or both kidneys, more often of the right alone.

In pyelitis the urine varies much in different cases, and at different times in the same case. The variation in the quantity of urine excreted, a decrease in the acute cases and an increase in the more chronic ones, which is usually present, may not be marked and is of course not distinctive of this condition. Pyuria is present, but it should be remembered that this does not always mean a densely turbid urine with a large amount of sediment. In fact, there may be at most but a very slight turbidity and the sediment correspondingly small. It is only a careful microscopic examination that may indicate the condition present. Another circumstance which is often overlooked is that when pus is collecting in considerable quantities in the pelvis of the kidney in pyelitis there temporarily may be no pyuria provided the condition is unilateral. This is explained by a blocking of the ureter of this kidney for the time by pus, or a calculus or a kink in the ureter. For this reason repeated examinations of the urine may be necessary to reach a diagnosis.

Will an examination of the urine in a case of pyuria show whether the pus is from the kidney pelvis rather than from some other part of the urinary tract? I have been interested in this for some years and do not agree with those who say that this cannot be done. I think that under favorable circumstances it is possible to do so. If there is pyelitis without cystitis this can be diagnosed fairly accurately from the urine. If cystitis is present, it is more difficult to determine whether the pelvis is also involved. There is perhaps no one thing alone in the urine which tells that we are dealing with a pyelitis, but there are several factors which combined may indicate the pelvis as the source of the trouble. The urine in pyelitis is most often acid, though not always. Fewer epithelial cells are present in

pyelitis than in cystitis, while the so-called tailed epithelial cells are relatively more numerous in pyelitis. I have found red blood corpuscles in small numbers more often in pyelitis. Together with the pus there are liable to be casts of the larger collecting tubules of the kidney, at times containing pus cells, which aid in confirming the diagnosis of pyelitis together with some involvement of the kidney parenchyma. I think that in the great majority of cases the diagnosis can be made from the urine. In those exceptional chronic cases in which the urine is constantly alkaline, not due to drugs, but where in spite of this the examination of the sediment is sufficient to indicate the condition to be pyelitis, and a cystitis can be excluded, the presence of a phosphatic calculus in the pelvis may be diagnosed with a fair degree of certainty. Crystals of acid salts in a freshly voided urine giving evidences of pyelitis may indicate a calcium oxalate or urate calculus. As already mentioned, when a cystitis is present the diagnosis of an accompanying pyelitis from the urine examination is more difficult. It then becomes largely a matter of finding the large tube casts with pus cells which show the involvement of the tubules in the pyramids of the kidney as an extension of a suppurative pyelitis. In the female it is often essential to obtain a catheterized specimen of urine, and in all cases where a bacteriological examination is to be made it should be obtained and kept under aseptic conditions and examined while perfectly fresh.

In considering the etiology of pyelitis we may disregard those cases due to drugs or to the toxins of the acute infections, as the cause is here evident and relatively unimportant except as a predisposing factor to bacterial growth. My own results in the bacteriological examination of the urine in pyelitis agree with those obtained by others, that the condition is practically always due to bacteria, that the offending bacterium whatever its nature is generally present in pure culture, and that the colon bacillus is the organism present most often. The other bacteria which are less often found as the cause of pyelitis are, however, of importance as some of these, like the B. proteus vulgaris and certain staphylococci, are capable of decomposing urine and

in this way are active factors in the production of phosphatic calculi in the pelvis. In females the colon bacillus is by far the most common invader according to all reports (Von Albeck, Lenhartz, Brown, Rovsing). In males (Rovsing) it appears to be much less common. And here mention should be made of the comparative frequency with which the tubercle bacillus is the cause of pyuria from the involvement of the pelvis and lower portion of the kidney substance. Any pyelitis with an acid urine, which cannot be satisfactorily explained by other bacteria present, should be considered quite possibly tubercular and a careful microscopic search made for the tubercle bacillus. It is as a rule not easily found and repeated examinations may be necessary, and often animal inoculation, before it can be demonstrated.

The presence of such common bacteria as the colon bacillus and the staphylococci as the active cause of pyelitis brings up the question of how they invade the pelvis of the kidney. I shall take but little time in discussing the route by which they reach the kidney. It has been well established that bacteria reach the kidney frequently through the blood, also by means of the ureter, and through the lymphatics, at times apparently coming directly through the wall of the colon to the kidney. The ascending route through the ureters is probably much less common than formerly believed. Even when pyelitis is secondary to a chronic cystitis the route may or may not be by means of the lumen of the ureter. This question of the route of invasion, though of scientific interest, does not appear to me to be of as much importance as are the underlying causes which favor infection of the kidney by the colon bacillus and other organisms. Under normal conditions the urine is germicidal and the epithelial lining of the urinary tract from the pelvis down is resistant to bacteria. The predisposing factors of pyelitis are such as cause here a local lessening of resistance to bacterial invasion or such as cause an increased virulence of the organism. Of these the former seems the more important. The most common appears to be some anatomical condition which interferes with the outflow of the urine, and particularly when

this interference has already been followed by the development of a cystitis. I need only to mention this as it is so well recognized as seen in cases of hypertrophied prostate, strictures, etc. The movable kidney with its passive congestion and often compression of its ureters with resulting slight hydronephrosis is not so often considered in this connection as it deserves. There are numerous other factors, only two of which I will mention. The first of these is constipation. As shown by Posner and Lewin the intestinal wall becomes pervious to bacteria in various pathological conditions, and such lesions of the mucous membranes as may occur in constipation may allow bacteria to enter the blood stream. Also the virulence of the colon bacillus is increased in diarrhea and other intestinal diseases. Digestive disorders thus may play a somewhat important part as accessory factors in the producing of pyelitis. Another factor is pregnancy including the puerperium. This is always mentioned as one of the common conditions predisposing to pyelitis but is forgotten until impressed upon one by meeting striking examples of it. I happen to have seen several of these cases occurring during the puerperium, mistaken for a late puerperal sepsis, and the error is such a natural one and the symptoms so confusing that it seems worth while to emphasize the condition here, although I have already referred to it. The patient may have done well following delivery for a few days but the chart often shows a slight rise of temperature during this time. After a week or ten days or more there is a chill, sudden rise of temperature to 103 or 105 with rapid pulse, all the evidences of an acute infection often with almost no localized symptoms. A characteristic finding in these cases is the lack of tenderness in the region of the uterus. Nor do the lochia give evidence of uterine infection. A catheterized specimen of urine at once shows the pyuria and careful palpation on the right side over the kidney will elicit slight tenderness there. This most often involves the right kidney. Opitz in 63 cases found. it limited to the right side in 66 per cent.; while Ward found the right side alone involved in 55 per cent. of 187 cases, in IO per cent. the left side alone, and in 35 per cent. both sides