closed by a sphincter which may retain the urine and afterwards voluntarily expel it.

It was observed in a case of congenital union of the ureters to the intestine, reported by Richardson (36), that a child who lived 17 years always urinated by the anus. Moreover a whole class of animals (the birds) have a single cloaca which receives both

urine and feces. Besides it is demonstrated that the presence of urine in the rectum does not cause uremia, nor does the rectum undergo grave changes: hypertrophy of the connective tissue and the follicles only being observed.

This subject is open for study. There are many questions to be cleared up. The most important is that which treats of ascending infection of the kidney by micro-organisms coming from the intestine. This infection, which some think always inevitable

whatever portion of the ureter may be infixed, seems to others avoidable if so much of the uretero-vesical opening of the orifice is respected as may be provided with a sphincter destined to protect the kidney. And indeed the experiments of Novaro in dogs, the clinical case of Chaput, that of Maydl, as well as that of Trendelenburg, demonstrate that infection of the kidney may be avoided. For if we analyze the observations of experimenters who have secured only negative results, we shall find that their attempts failed either immediately because the ureter did not adhere, in which case death followed during the first days from urinary infiltration of the peritoneal cavity, or rather because the kidney had been infected, in which case death followed at an earlier or later period. In cases in the last category, at the autopsy the experimenters in most cases found stenosis of the uretero-intestinal opening and the corresponding kidney in a state of hydronephrosis. For, as we know by the experiments of Albarran and Robinson, and the clinic confirms it, hydronephrosis is an inevitable result of stenosis of the ureter, while the sudden and complete obliteration of this canal is followed by atrophy of the kidney.

Resting upon these experiments it is not possible to assert with certainty that the cause of failure depended upon an infection ascending from the intestines to the kidneys; still less can we conclude that this infection is inevitable and that the operation should be condemned by experience itself. On the contrary, the infectious nephritis constantly observed by some experimenters, is probably due to a predisposition caused by a hydronephrosis arising from a defective graft. Indeed should some able operator succeed by the aid of sutures in fixing the thin ureter into the intestine, so that the urine does not pass into the peritoneal cavity, it is difficult to believe that the ureteral end will not be somewhat contracted by a point of suture. And let us suppose that the operation is perfect in the highest degree, must not we fear that the tissues surrounding the ureter, as a consequence of the disposition necessarily resulting from the suture inserted, may contract the uretero-intestinal opening at the time of cicatricial condensation. Some experimenters have not occupied themselves solely with making the graft; they have simply buried the ureter in a button hole in the intestinal wall, fixing it by points of suture, or else they have introduced into the extremity of the ureteral open

ing small tubes or foreign bodies badly adapted to this purpose. Doubtless these methods were responsible for immediate failure or tardy stenosis. But Novaro, who has occupied himself in obtaining an ample and large anastomosis by his method of sutures in V, has succeeded in dogs where the ureter is very small and where it is especially difficult to keep an open orifice.

Chaput, who operated on a woman with a sound kidney with a largely dilated ureter, as he tells us in his memoir, obtained a complete success which has lasted already three years. These results, although small in number, demonstrate that infection of the kidney is not inevitable. Indeed other organs have their excretory ducts which empty into the intestine, as the liver, the pancreas, the glands which carpet the gastro-intestinal tube and which are not infected according to the order established by nature. The uterus and the ovaries have their excretory tubes which open into non-aseptic cavities. In order to explain this phenomenon they have recourse to the antiseptic power which is possessed by the glandular secretion of some of these organs. In others they explain the fact by the valve action of the sphincter, of contractile fibres destined perhaps to hinder the exterior agents. of infection reaching the organ. Certainly nature which fashioned them thus has given a means of defense.

For myself, I am convinced that the current of the fluid secreted by an organ, when it has free course, is capable of maintaining the excreting canal intact and of defending it from infecting agents which may be introduced from without. It is thus with the ureter-the descending current of urine washes and cleanses it continually. Even the pressure of the secretion is a means of defense. Besides the uretero-vesical opening has a sphincter which is doubtless destined to protect the kidney, as Tuffier demonstrated; but it seems to me that the ureteral sphincter ought to be employed less in hindering the ascent of infection, for which the descending current of the urine and also the ejaculations are sufficient, than by preference in giving to the bladder the function of a true reservoir, whatever position the human body may take, and, in acting as a valve which opposes the regurgitation of the urine when the bladder is full. For ascending infectious ureteritis, as Chaput observed, cannot occur without the aid of two very different mechanisms. In one the urine becomes a vehicle for the microbes when an opposing obstacle slows the current in the

ureters. In the other, infection may occur by direct propagation on the part of the lymphatic vesels, or indirectly layer by layer across the tissues.

The second factor of failure may be eliminated if, at the moment of the operation, the intestine is sound. Ascending inflammation of the lymphatics or across the tissues would not arise except under the irritating influence of the urine, but experiments have proven the harmlessness of the urine in the lower portion of the intestine, and comparative anatomy demonstrates the same. thing in birds, where the urine is turned into a single cloaca, besides many observations in human beings of vesico-vaginal fistulas being tolerated many years by the intestinal mucous membrane.

In view of all these reasons we have a right to think that ascending infection of the ureter is not solely due to defective conditions in the course of the urine. Hence arises the duty of directing researches upon the method to be adopted to successfully and easily make the graft, the ureter being open and accessible with the least production of connective tissue, and with the aim of preventing later stenosis. Led by this idea I have constructed a button to replace the sutures and be expelled by the anus when cicatrization is complete.

FIRST EXPERIMENT.-My first experiment was published in the Periodique Italien de Chirurgie of Prof. Durante (Policlinico, vol. 11, C. fasc. 10, 1895); to the Tenth Congress of the Italian Society of Surgery at Rome. I reported the first two clinical applications of my button, exhibiting there the buttons in question, as well as the anatomical specimens obtained from the animals.

The first anatomical specimen shows a bilateral graft, obtained in a hunting dog of medium size, by the abdominal route by the aid of button No. 2. (The specimens are preserved in alcohol.) The ureters are not distended; the tunics of the left are a little larger in front of the opening into the rectum, i. c., at 9 cm. from the anus; it is fixed for a certain length upon the intestinal serous membrane by connective tissues. The orifices of the opening into the intestine are perfectly visible and accessible; the right is surrounded by a little papilla like a valve and is found 14 cm. from the anus. The kidneys macroscopically show no grave alterations. The right weighs 70 grammes, the left 75 grammes. The

capsule can be detached with ease. Upon section there is no hydronephrosis. The renal substance seems normal; it is only to be remarked that the cavity of the pelvis is slightly dilated a little on the right side which was grafted last. Perhaps the slight dilatation of the pelvic cavity has its importance, but I am not able to tell whether it indicates the beginning of an affection with a ten

GRAFT OF RIGHT URETER UPON THE RECTUM.

FIG. 3 AND 4 of this plate showing the Boari Button open and closed respectively.

dency to increase, or rather the resolution of the consequences of the operation. The last hypothesis seems to me the more probable because it is more manifest on the side last operated on. The urine cannot have ever been stagnant, otherwise there would have been hydronephrosis. Perhaps the slight degree of dilatation is due to the fact that the new orifice of the opening lacks the sphincter and the increase of pressure in the rectum ought to be prop