tended with clear fluid having the appearance of urine. No trace of true renal structure could be found. The parenchyma had been absorbed and was represented by a wall of fibrous tissue containing fluid (hydronephrosis.) A probe passed down the ureter could not be made to enter the bowel. The right kidney was much enlarged, being as large as the organ in an adult. The capsule adhered to the cortical substance, which was soft and friable and contained many small cysts. The medullary substance was much softer than natural. The descriptions terminate with the remark, "that, setting aside the difficulties and dangers of the operation, it would appear that a permanent and direct communication between the ureter and the bowel is in itself a fatal lesion."

For a number of years the operation seems not to have been repeated. Then Kuster, in 1891, after removing the bladder for cancer, implanted both ureters into the rectum, but the patient only lived five days. The autopsy revealed peritonitis and kidney infection. Chaput, in 1892, reports one case of unilateral implantation with success, and another case of bilateral with death from uremia. This last case was divided into two stages, death following the same day after the second ureter was implanted. Other operators, viz.: Rein, 1894; Duplay, 1894; Trendelenburg, 1895; Schnitzler, 1898; Truetta, 1899; Krause, 1899; Wood, 1899, and Martin, 1900, reported cases of bilateral implantations, all with fatal results. Successful results were reported after bilateral implantations by Chalot, 1896; Peters, 1899; Krause, 1899; Fowler, 1896; Tuffier and Dujarnier, 1896, and Beck, 1899.

Peterson gives the mortality of bilateral primary implantation of the unguarded ureters at 44 per cent. The ultimate mortality must be much higher, as ascending pyelitis is present in the majority of cases, and, although the immediate effect can be recovered from, the remote results will be disastrous. Thus Wood, in 1899, showed before the Brooklyn Pathological Society the kidneys of a patient suffering from exstrophy of the bladder, both of whose ureters two months previously had been implanted into the rectum by Fowler's method. There was marked pyelo-nephritis.

The list of recoveries is, however, sufficiently remarkable to

show that under certain conditions, which we do not properly understand, pyelitis does not necessarily follow the implantation of an unguarded ureter in the bowel. Or, if it be conceded that pyelitis does occur, that it is possible for the kidney to recover from the infection. In one of Peterson's experiments in dogs the animal that lived the longest showed indisputable evidences of a healed pyelo-nephritis. Contracted kidneys had resulted on both sides. (Journal Americal Medical Association, March 16, 1901, page 738.) A resumé of the successful cases is of great interest. Chalot. Cancer of the uterus; bilateral uretero-rectal anastomosis; recovery; living one year after the operation. Method of operating.-Author's specially constructed tube.

Fowler. Exstrophy of the bladder; bilateral uretero-rectal anastomosis; recovery; well three and a half years after the operation. Method.-Flap method; the ends of the ureters being applied close to the mucous membrane of the bowel, which was doubled up to form a recess. This method was original.

Tuffier and Dujarnier. Cancer of bladder; excision; bilateral uretero-rectal anastomosis; death seven month later; no autopsy. This case had leaking of urine from suprapubic fistula; therefore, it can not strictly be classed amongst the other cases.

Peters. Exstrophy of the bladder; bilateral extra-peritoneal uretero-rectal anastomosis; recovery.

Krause. Removal of bladder for cancer; implantation of both ureters into sigmoid; recovery.

Beck. Tuberculosis of bladder; bilateral implantation of both ureters into the sigmoid; recovery.

The contrast between the results obtained by the axial implantations into the bowel and those reported by Maydl in 1894 of implantation of the trigonal orifices is very striking. Maydl utilized the principle suggested by Tuffier of transplanting the vesical orifices of the ureters into the rectum to prevent ascending infection and stricture, and reported two cases with two recoveries. His method was to dissect out the two ureteral orifices, together with an oval area of the trigone, the edge of the area being about 1 c.m. away from the orifices. The ureters were then liberated from

the tissue in which they were imbedded and lifted up with the trigone attached. The rest of the bladder wall was then excised. The sigmoid flexure was pulled out of the wound, an opening made in its convex border, and the trigone and ureters placed inside the bowel. The opening in the bowel was then closed with two layers of sutures, one uniting mucous membrane, the other muscular and serious coats. That the preservation of the trigonal orifices of the ureters prevents ascending infection passing from the rectum to the kidney has been proved beyond denial. Thus in 1899, only seven years after the first operation was performed, Maydl collected twenty operations, including his own, with only three deaths. This makes a mortality of a little over 13 per cent, which, considering the difficulties encountered and the delicate nature of the procedures involved, is little short of remarkable.

Still, in spite of the facts observed clinically on man, experimentation on dogs showed the same disastrous results notwithstanding the precautions taken in operating. Thus Matas (Journal American Medical Association, July 29, 1899, page 260) performed a typical Maydl's operation on two dogs with a fatality in both, one dying in twenty-four, the other in thirty-six hours. In neither was there leaking of either urine or feces. The grafts were perfect. But in both there was evidence of septic peritonitis caused, so Matas thought, by infection from the bowel during the operation. In the dog that survived thirty-six hours there were evidences of a very intense hyperemia of the kidneys.

Connell (Journal American Medical Association, March 9, 1901, page 635) also failed in his experiments on three dogs with the typical Maydl operation (experiments 20, 21, and 24, page 666). In these cases death resulted with evidences of peritonitis and ascending infection of the kidney. He remarks that the cause of the peritonitis is difficult to arrive at.

Peterson's experiments on dogs show better results. By improving the technique and operating so as to preserve the vesical arteries supplying the trigonal flap, he managed to secure five recoveries out of twenty-one dogs operated on. His results go to show that in dogs the vesical orifice of the ureter does protect

against intestinal infection. (Journal American Medical Association, March 23, 1901, page 808.) To quote his own conclusions (page 811, loc. cit.): "Summarizing the results of these five experiments, we find that in one case the ureter was occluded from faulty technique with the formation of an atrophic kidney. Of the four remaining cases, one had a non-infected kidney where the trigone was implanted intact, and a pyelo-nephritis on the other side where a uretero-rectal anastomosis by the old method had been made. One died in forty-four days from active pyelo-nephritis, where the mucosa over the ureteral orifices had been accidentally removed. One lived two months without signs of infection in the kidneys. One had pyelo-nephritis after eight days, where the mucosa was removed from the ureteral orifices, but no signs of infection in the other kidney where the trigone was implanted intact."

The almost uniformly fatal results following implantation of the human ureters, unguarded by the vesical orifice, into the intestinal canal shows that the kidneys are just as susceptible to ascending infection as those of the lower animals. Perhaps, in the future, we may be able to implant an unguarded ureter into the intestine with safety; but, at present, the operation is so sure to be attended sooner or later by destruction of the kidney that it almost amounts to a nephrectomy.

The technique of the various methods of ureteral implantations has received much study, and from the experimental work of the investigators we have learned much concerning the best way to prevent urinary and fecal extravasation. The names of Krynski, Fowler, Martin, Van Hook, Beck, and Peterson are inseparably connected with the literature of the subject. The consensus of opinion seems to be in favor of an oblique insertion of the ureters into the intestine, the ureters coursing for some distance between the muscular and submucous layers before entering the lumen of the gut. To affect this the muscular and serous layers are thrown up in form of a flap (Krynski, Beck), or dissected back like skin flaps (Fowler, Martin), an opening being made in the mucous membrane at the most protected place, through which the ureter

is inserted, after which the muscular and serous coats are sutured back into place, and if necessary the coats of the intestine puckered over the line of implantation by Lembert's sutures.

The method of holding the ureter in place is deserving of the greatest consideration. In the main, the method of Van Hook is

the best. This is as follows: A thread is armed with two needles, one at each end. Each needle is thrust through the wall of the ureter from within outwards and the ends of the thread drawn tight, thus controlling the ureter by a loop. The gut is now opened in the most approved manner, and the needles are passed in at the opening and made to penetrate the wall close to one another from