Conclusions Regarding the Use of Drainage Tubes and Liga- tures. and the Possibilities of Skin Disinfection based upon Bacteriological Investigations BY HUNTER ROBB, M.D. Associate in Gynecology to the Johns Hopkins Hospital, Baltimore. Md. KEPKL.NTED FUOM The American Journal of Obstetrics Vol. XXVI., No. 6, 1892 NEW YORK WILLIAM WOOD & COMPANY. PUBLISHERS 1892 CONCLUSIONS REGARDING THE USE OF DRAINAGE TUBES AND LIGATURES, AND THE POSSIBILITIES OF SKIN DISINFECTION BASED UPON BACTERIO- LOGICAL INVESTIGATIONS.1 The Use of Drainage Tubesand Ligatures.-Although to the use of drainage we have to attribute great improvement in abdominal surgery, the tendency of late, owing to our better knowledge of wound processes and the modes of wound infection, has been toward the abandonment of the practice. Drainage of cavities by means of tubes has been employed since the days of Hippocrates and Galen. In the works of Hippocrates the use of hollow pencils is mentioned for this purpose, and Galen speaks of leaden tubes in the same con- nection. In 1731 Heister, of Nuremberg,2 recommended wick drainage in large abdominal wounds until the discharge be- came perfectly healthy. Chassaignac, of Paris, in his work, " Traite pratique de la Suppuration et du Drainage,'1 pub- lished in 1859, describes the principles underlying the system of drainage. He it was who introduced the rubber tube, and 1 Read before the First International Congress of Gynecology and Ob- stetrics, held in Brussels, Belgium, September, 1892. ■iLaurentius Heister's "Surgery," Nuremberg, 1731, third German edi- tion, p. 88, chapter vii., section 3. 2 ROBB: DRAINAGE TUBES AND LIGATURES, it is to him, therefore, that the credit of being the first to em- ploy drainage is generally given. The tube used during the past quarter of a century is the glass one, and to Koeberle, of Strassburg, belongs the honor of introducing drainage tubes of this material. The type devised by Koeberle has been materially modified from its original form, showing, by the many improvements which have been made in it, the ad- vances in abdominal surgery. Probably one of the first to see the necessity for these changes was Keith, of Edinburgh, to whom Koeberle had given two of his tubes. Keith found that by employing larger tubes and by the removal of the closed bottom he was enabled to obviate the tendency of the tube to become choked up with clots of blood and lymph. He states that if he had employed the drainage tube earlier his mortality in abdominaljBurgery would have been lessened by one-third. In 187G las'death rate at the Samaritan Hospital, London, was reduced to ten per cent-a reduction which he attributes entirely to the use of the drainage tube. In spite, however, of all this, the statement so frequently made that to the drainage tube alone, or almost entirely, is due this decrease in the mortality in abdominal surgery, cannot be allowed to pass without criticism. No one, we think, will dispute the fact that the improvements made in the technique of operations must also be taken into considera- tion. The dangers which we now know to exist with open abdominal wounds were until recently but imperfectly ap- preciated. Now that the conditions underlying the infection of wounds are so much better understood, and it is becom- ing recognized that the drainage tube too often permits the entrance of pathogenic bacteria, we can understand why its general use in abdominal surgery has been abandoned by many operators. There are, of course, circumstances even yet where the tube is indispensable, but we believe that the necessity for its use after abdominal operations will occur but rarely. Thus, in cases of pelvic abscess where it is impossible to remove entirely the diseased structures, where, for instance, pus is free in the pelvic cavity or is contained in a sac which can- not be taken out, it may still be well to drain. AND SKIN DISINFECTION. 3 Again, in rare instances bleeding may be so free that one fears to close the abdomen, and the tube might then be em- ployed for a short time in order that we may be better able to watch the extent of the hemorrhage, so that, if the bleed- ing continue, we can reopen the abdomen and endeavor anew to control the bleeding points. Such a procedure will offer an additional safeguard, since the pulse cannot always be de- pended upon to give an exact indication as to the amount of bleeding that is going on. We are, however, of the opinion that, where the technique has not been careless, the necessity for the use of the tube even for this purpose will not often arise. A small amount of bleeding (oozing) is, as a rule, of little consequence, and the peritoneum is able to absorb the fluid, which is not a source of danger unless infected. Three years ago we held the view that the drainage tube was of value in almost every case of abdominal section; but after a careful bacteriological analysis of over one hundred cases, and the discontinuance of the practice of drainage of any kind in more than one hundred abdominal sections, we were led to agree with the conclusions so admirably formu- lated by Prof. Welch in an address delivered before the Clinical Society of Maryland in the fall of last year.1 Speak- ing of wounds in general, Dr. Welch raises the following objections to the insertion of drainage tubes : "1. They tend to remove bacteria which may get into a wound from the bactericidal influence of the tissues and ani- mal juices. 2. Bacteria may travel by continuous growth or in other ways down the sides of a drainage tube, and so penetrate into a wound which they otherwise would not en- ter. We have repeatedly been able to demonstrate this mode of entrance into a wound of the white staphylococcus found so commonly in the epidermis. The danger of leaving any part of a drainage tube exposed to the air is too evident to require mention. 3. The changing of dressings necessitated by the presence of drainage tubes increases in proportion to its frequency the chances of accidental infection. 4. The drainage tube keeps asunder tissues which might otherwise immediately unite. 5. Its presence as a foreign body is an irritant and increases exudation. 6. The withdrawal of tubes 1 Welch, Maryland Medical Journal, 1891. 4 kobb: drainage tubes and ligatures, left for any considerable time in wounds breaks up forming granulations-a circumstance which both prolongs the process of repair and opens the way for infection. Granulation tis- sue is an obstacle to the invasion of pathogenic bacteria from the surface, as has been proven by experiment. 7. After removal of the tube there is left a track prone to suppurate and often slow in healing." To these Prof. Halsted has added an eighth : " That tissues which have been exposed to the drainage tube are suffering from an insult which more or less impairs their vitality and hence their ability to destroy or inhibit micro-organisms." If an abdominal wound becomes infected subsequent to an operation it is generally thought that this may arise from micro-organisms already present, it may be in a pelvic ab- scess or in the secretions in the uterine adnexa ; and we have heard it frequently stated by an operator that it was little wonder that the wound became infected when pus existed previous to the operation. Undoubtedly this mode of wound infection may occur, but it should be remembered that in a very large proportion of pyo-salpinx cases the pus is sterile, and if any organisms are present they are frequently dead. This has been proven many times by examination of smear cover-glass preparations and the study of cultures made at the operation. If gonococci are present there is no evidence that they are capable of infecting wounds. Unless bacteriological examinations are made of such secretions or pus accumula- tions, it is impossible to feel sure that an infection which has followed an operation has come from within. In instances in which an examination has shown the presence of living or- ganisms, and the case has terminated fatally with the same organisms present in the wound and peritoneum, we can fairly assume that this was the way in which the infection occurred, bio reliance can be placed on most of the state- ments made with regard to the finding of gonococci in the secretions, if the demonstration is only a microscopical one, as in cover-slip preparations ordinary staphylococci may be enclosed within leucocytes and may so closely resemble gono- cocci that the two are morphologically almost, if not quite, indistinguishable ; and pus from a gonococcus infection, even if allowed to escape into the pelvic cavity, does not always AND SKIN DISINFECTION. 5 set up inflammatory changes, although an occasional case of gonococcus peritonitis has been reported. But cases of ovariotomy and hysterectomy which are free from pyogenic organisms prior to the operation frequently become infected, and such an infection, when not due to the skin coccus-about which we shall speak later-must be looked upon as the direct result of some fault in the technique. In hysteromyomectomy one has always to think of the danger of infection by way of the cervical canal and vagina. Dbderlein,1 in his recent work on the vaginal secretions, has shown that in eleven per cent of the women with pathologi- cal vaginal secretions virulent streptococci were present. In order to determine what part the drainage tube plays in the origin of infection in abdominal wounds after opera- tion, we have undertaken a series of bacteriological examina- tions of the secretions which accumulate in the tube, as well as those which saturate the gauze plug that is placed within it. With the assistance of Dr. A. A. Ghriskey, formerly as- sistant resident gynecologist to the Johns Hopkins Hospital, these experiments were carried out in the gynecological wards of Prof. Kelly, to whom we are greatly indebted for the opportunities afforded. The bacteriological examinations were made in the Pathological Laboratory of the Johns Hop- kins University and Hospital under the supervision of Prof. Welch. In a series of cases where the drainage tube was»employed we were able to demonstrate the impossibility of maintaining a perfectly aseptic condition, in spite of the most painstaking precautions. The unfavorable results from the use of the drainage tube are in many instances owing to the introduc- tion of bacteria by the operator himself in the act of dress- ing the tube, by his assistant, or by the nurse to whom the duty is delegated. Reasoning from our knowledge of the distribution of bacteria, it is too much to expect that a drain- age tube could be placed and kept in an open cavity without in some way often becoming contaminated, especially if it be now and again exposed to the air. In making our experi- ments the following technique was observed : 1 " Das Scheidensekret und seine Bedeutung fur das Puerperalfieber." Von Dr. Albert Dbderlein. Leipzig, 1892. 6 ROBB : DRAINAGE TUBES AND LIGATURES. On dressing the abdomen immediately after the opera- tion a piece of rubber was taken, a tritie thicker than the rubber-dam used by dentists. This had previously been sterilized by being allowed to soak in a watery solution of bichloride of mercury (1 :500) for three hours, and after- ward had been kept in sterilized salt solution. In the centre of the piece, which was large enough to cover the abdomen from flank to flank and from the symphysis pubis to just be- low the ribs, a slit was made through which the drainage tube was allowed to protrude; over the rubber was placed a piece of cotton previously sterilized by steam and about the size of the closed fist, and over this the ends of the rubber-dam were folded. In this way the possibility of the entrance of septic material from without was reduced to a minimum. The cot- ton was sufficient to absorb any fluid which might come through the tube by the capillary action of the sterilized gauze placed within it. Over this dressing again was laid enough sterilized cotton to protect the abdomen and to allow the bandage to be neatly adjusted. We found that the most efficient ma- terial to drain the tube was ordinary cheese-cloth cut into strips about two centimetres wide and forty centimetres long, which had previously been rolled and placed in glass tubes and sterilized by steam. At the dressings one of these strips was removed from its glass tube and carried to the bottom of the drainage tube, and if the strip of gauze were too long the portion projecting above the top was cut off with steril- ized scissors. In the subsequent dressings, after removing the gauze from the tube, we made use of small sterilized cotton pledgets rolled into balls, just large enough to fit the calibre of the tube. These were carried to the bottom by Dr. Kelly's ster- ilized tube forceps, and by them any fluid that might have collected was soaked up. The pledgets were removed from the drainage-tube forceps by sterilized dissecting forceps held in the other hand. In this way all fluids could be suc- cessfully removed without the use of a syringe-an instru- ment which readily becomes contaminated. In handling these materials thin rubber gloves were worn which had been soaked previously for five minutes in a watery solution of the bichloride of mercury (1:500). They were put on and AND SKIN DISINFECTION. 7 washed off in sterilized salt solution just before the tube was dressed. The instruments were sterilized by means of steam, and at no time did the tinkers come in contact with the tube. No antiseptics were introduced into the tube. At each dressing four roll or Esmarch agar cultures, and in addition two smear cover-slip preparations, were made. Two agar-agar Esmarch and two four-per-cent glycerin-agar- agar tubes were inoculated, one of each kind from the fluid at the bottom of the gauze plug, and the remaining two from what was obtained by scraping a platinum needle along the side of the plug. Two cultures were made also after intro- ducing the platinum needle down into the pelvis to the bot- tom of the drainage tube. Two smear cover-slip preparations of the secretions from each place were also made at once. These were stained with methylene blue or gentian violet, and careful search made for bacteria. The first series of cases that we examined convinced us that in but few instances are the secretions in the drainage tube free from organisms. The second series clearly demonstrated that our first observations were correct, and made clear the dangers of the drainage tube. The total number of laparatomy cases observed in this series was forty-five, and in no less than thirty-one, or sixty-nine per cent of the whole, the presence of organisms was demon- strated, while in only fourteen were the results negative. The frequency of the occurrence of the different cocci was as fol- lows: 1. The staphylococcus albus in nineteen cases. 2. The staphylococcus aureus in five cases. 3. The bacillus coli communis in six cases. 4. The streptococcus pyogenes in only three cases-twice in combination with the albus, and once alone. In three of the aureus cases the organism was present in the diseased focus before the operation, while in the other two infection from outside could not be excluded. Of the streptococcus series the first was a case of ovarian abscess which was found at the time of operation to be full of streptococci; and it is quite possible that the remaining two cases, which occurred soon after, were due to infection from this patient, though the mode of conveyance could not be traced. The clinical history of the cases belonging to the different series presents a striking contrast; for while in all 8 ROBB : DRAINAGE TUBES AND LIGATURES, the staphylococcus aureus and streptococcus cases there was suppuration and the patients were seriously ill, in the white staphylococcus cases, on the contrary, the condition was gene- rally favorable and the healing of the wound was but little if at all delayed. The best results of all were obtained in those cases where this coccus occurred in small numbers and not until twenty-four or forty-eight hours after the opera- tion and after the first dressing. Even in the cases where it caused some trouble this was never of any real moment. We are led to believe that in some cases where there was fever without suppuration the fever was due to an albus infec- tion, and to the same agent may be imputed the stitch ab- scesses not infrequently met with, and in which we found this organism often in pure culture. Our observations point most certainly to the drainage tube as the vehicle of infec- tion, for it must have been along the tube that this coccus, which is found with great regularity in the epidermis, travelled to the bottom of the wound, and wherever there was a puru- lent discharge it was always found in the tissues that had been in contact with the tube. While, however, in the ma- jority of cases the comparative innocuousness of the albus seems to have been proven, the findings in one case of fatal hysteromyomectomy show how, under exceptional circum- stances, this usually innocent coccus may become virulent. At the autopsy the stump and the laparatomy wound were perfectly healthy, but there existed a volvulus of the ileum, and the peritoneum covering the twisted part of the gut, and at the same time adherent to the inner edge of the laparatomy wound, was found to be the seat of a fresh fibrino-purulent peritonitis. There was no trace of gangrene or any marked hemorrhagic condition. In the inflamed peritoneum a pure culture of the staphylococcus albus was found, and it was evident that the twist in the gut had interfered with the cir- culation and nutrition just enough to afford a favorable soil for the development of the coccus, which under these cir- cumstances was virulent enough to produce a fatal issue. Skin Disinfection.-Besides these experiments bearing on the drainage tube, Dr. Ghriskey and myself, under Prof. Welch's supervision, made a series of bacteriological ex- AND SKIN DISINFECTION. 9 aminations of scrapings from the skin. The cultures were taken from the hands and from the abdominal surface. These experiments on the skin of the abdomen were made in the following way : The skin in the median line just be- low the umbilicus, for a distance of six by six centimetres, was first cleansed with absolute alcohol applied on sterilized absorbent cotton; then, a fold of skin being held firmly be- tween the thumb and second finger of the left hand, the sur- face was scraped with a sterilized knife blade. In most instances the scraping was deep enough to produce slight ooz- ing, and the deeper portions of the loosened skin were im- mediately planted in nutrient agar-agar and gelatin tubes, three tubes of each being employed in every case. The scrapings were taken from patients under anesthesia, just before an examination of the pelvic organs. A record was always made as to any peculiarities of the skin, especially if an eruption or any scars were present. In nineteen out of twenty-five cases examined in this way the results were posi- tive, and the organisms were in most instances identical with those found in the secretions from the drainage tube. By far the most constant organism seen was a white sta- phylococcus, which has been differentiated and named by Prof. Welch1 the Staphylococcus epidermidls albus. It is found almost constantly in the epidermis, and, as he remarks, may be an attenuated form of the staphylococcus pyogenes albus. This coccus is sometimes found in the graver forms of suppurative inflammation, but in these cases is nearly always associated with some other pyogenic organism or has assumed the form of the typical staphylococcus pyogenes albus. We also made use of Fiirbringer's method of disinfection of the skin, and found, after precipitation of the mercury with an alkaline sulphide, this same white staphylococcus in the majority of cases in the scrapings from the epidermis. A series of forty-five consecutive examinations was made of stitches removed during and after operations on the abdo- men and perineum. The stitches were examined microsco- 1 Welch, "Conditions underlying the Infection of Wounds." Transac- tions of the Congress of American Physicians and Surgeons, vol. ii., 1891. 10 ROBB : DRAINAGE TUBES AND LIGATURES, pically and by culture methods, as were also the fluids along the line of the incision. The stitches were passed by means of disinfected needles and holders, with the avoidance of every possibility of external contamination through the skin after it had been disinfected by a method to be described subsequently, and which yields negative results from scrap- ings from the surface of the skin. In these investigations we were able to demonstrate the presence of the skin coccus with great regularity, thus confirming our previous work and showing that this organism is present in layers of the skin deeper than can be reached by existing methods of disin- fection, so that we can understand how external wounds subjected to the most rigid antiseptic treatment may become infected from the skin of the patient. Culture tubes of nu- trient agar or gelatin inoculated with scrapings from the skin, after thorough disinfection of its surface by the method to be described later, remained sterile. At the time of operation a silk ligature, sterilized by steam and proven by culture methods to be free from organisms, was carried through the superficial layers of the disinfected skin, and in some instances through skin, muscle, and perito- neum, after the incision had been made through the abdomi- nal walls. From these, roll or plate cultures were immedi- ately made, as well as cover-slip preparations, and in nearly every case the white staphylococcus was demonstrable-often in considerable numbers. By this ready method of making cultures of stitches di- rectly from the deeper layers of the epidermis, the presence of the white staphylococcus often in pure culture has been re- peatedly demonstrated after complete superficial disinfection, in parts of the epidermis too deeply situated to be acted upon by any existing methods of disinfection. In the stitches removed after the operation similar results were obtained, the organisms being often enclosed in the leu- cocytes, with polymorphous nuclei not only where a stitch ab- scess had formed, but where there was macroscopically not a trace of suppuration or visible reaction about the seat of the stitch. From a large number of observations Dr. Welch concludes AND SKIN DISINFECTION. 11 that this coccus may be regarded as a nearly if not quite con- stant inhabitant of the epidermis. His reason for making a distinction between it and the staphylococcus pyogenes albus is based upon some cultural differences and the non-virulent character of the former, which possesses such feeble pyogenic powers (as shown by its behavior in wounds as well as by inoculation experiments on rabbits) that the designation staphylococcus pyogenes albus would not seem to be quite appropriate. The number of bacteria present depends upon several con- ditions. They are, however, always more abundant where the drainage tube is employed, for in these cases there is an increased amount of secretion, both immediately around the tube and on the sutures nearest to it, especially where the tis- sues have been unduly constricted and their resistance to the growth of organisms has been thereby diminished. The number of bacteria is also influenced by the form of suture material employed. Catgut sutures offer less resist- ance to bacterial invasion, and furnish a soil more favorable for bacterial increase, than silkworm gut, silk, or silver wire sutures. The knot of the catgut suture in skin wounds was found to be especially rich in organisms. Silkworm gut is the most resistant and harmless suture material, in our expe- rience, that has as yet been used. Its surface, smooth, com- pact, and without interstices, does not afford a good nidus for bacterial development; and although we have found many bacteria on them, yet they were fewer in number than those on either silk or catgut. The amount of secretion present <>n the silkworm gut is at times very slight, rendering it an ex- ceedingly difficult task to procure a sufficient quantity for a microscopical examination. If the silkworm-gut sutures are too dry for microscopical examination, one might be tempted to think that they were entirely free from bacteria. This, however, we have shown is not the case, for if bouillon cul- tures or slanting agar cultures are made of this dry ligature we obtain luxuriant growths of organisms. Again, silkworm gut does not produce the same constriction as either silk or silver wire, but acts more like a splint to the tissues. In order to increase the resistance of catgut and silk sutures to bac- terial invasion, we have recently used them paraffined. This 12 robb: drainage tubes and ligatures, is done by drawing the sterilized sutures through sterilized liquefied paraffin and allowing them to harden in absolute alcohol; but as to the results we are unable as yet to make any definite statement. We desire especially to emphasize the importance of ascer- taining the kind of bacteria present in the wound or on the stitches. For example, in three instances the streptococcus pyogenes was found both on the sutures removed and in the secretions in the incision. T wo of these cases occupied ad- joining beds; well marked suppurative changes with sys- temic disturbances followed in each case. We immediately isolated them, thus, as we believe, preventing the infection of other cases. In these patients the drainage tube was not used, and this fact, in all probability, prevented a streptococ- cus peritonitis. Although we cannot give here a full report of this work, which has been already published,1 we shall give a brief an- alysis of the cases: The Staphylococcus pyogenes aureus and the Staphylococcus epider- midis albus associated on one or two stitches 3 cases. Staphylococcus pyogenes aureus on every suture associated with the albus 2 " Staphylococcus gilvus 1 case. Staphylococcus gilvus and albus 2 cases. Streptococcus pyogenes with Staphylococcus epidermidis albus. ... 1 case. Streptococcus pyogenes . 3 cases. Staphylococcus epidermidis albus in pure culture 33 " Total number of cases 45 Six rabbits wrere inoculated with the staphylococcus epi- dermidis albus (skin coccus) from six different sources with- out positive results. The rabbits invariably looked and seemed ill, but they recovered. The dose varied from 0.5 to 1.5 cubic centimetres of bouillon culture twenty-four hours old. One rabbit, indeed, died after twro weeks, but no evi- dences of infection were found at the autopsy and no growths occurred in the Esmarch tubes made from the organs. From these observations we felt justified in drawing the following conclusions: A wound of the skin, at some stage of its existence, nearly 1 " The Bacteria in Wounds and Skin Stitches." By Hunter Robb and A. A. Ghriskey. Johns Hopkins Bulletin, No. 21, April, 1892. AND SKIN DISINFECTION. 13 always contains bacteria. They occur both on the stitches and in the secretions. The character and virulence of the organism present will, of course, influence the process of healing. The body temperature is invariably elevated if the bacte- ria are virulent; and, indeed, in cases where many of the less virulent organisms are found there is some fever. Different suture materials offer different opportunities for bacterial development, and the catgut sutures would seem to be the best adapted to their growth. In the event of the presence of the streptococcus pyogenes or staphylococcus pyogenes aureus such cases should be isolated, as far as possi- ble, to prevent the infection of others. Undue constriction of the tissues by ligatures must be avoided if they are to be expected to resist bacterial inva- sion. We have no absolutely sure method of rendering the field of operation entirely free from organisms, owing especially to the impossibility of destroying the bacteria in the deeper layers of the epidermis or in its glandular appendages. The staphylococcus epidermidis albus (skin coccus) is found in the skin with such regularity that this situation may, for all prac- tical purposes, be regarded as its natural habitat. Our conclusions regarding the dangers of drainage tubes have already been stated and need not here be repeated. Before bringing this paper to a close I beg leave to direct attention for a moment to the method of hand disinfection which we have found to be the best. Since the institution of bacteriological control as a test of the efficiency of surgi- cal technique, many methods before believed to be reliable have been proven to be faulty. Thus, the germicidal effect of the solutions-of corrosive sublimate has been shown to be, under ordinary conditions, less than was formerly supposed, what was interpreted by former observers as evidence of de- struction of bacteria often amounting only to an inhibition of growth. Geppert first drew attention to the fact that when, after disinfection experiments, the mercury is precipitated by ammonium sulphide, corrosive sublimate is a less efficient germicide than has been believed. Fiirbringer's method has in the course of our work been 14 robb: drainage tubes and ligatures, weighed in the balance and found to accomplish less than is claimed for it by its originator. Our best results, although not perfectly satisfactory, have been obtained by the follow- ing method: 1. The nails are kept short and clean. 2. The hands are washed thoroughly for ten minutes with soap and water, the water being as hot as can be comfortably borne, and being frequently changed. A brush, sterilized by steam, is used, and any excess of soap is washed off with water. 3. The hands are immersed from one to two minutes in a warm saturated solution of permanganate of potash. 4. They are then placed in a warm saturated solution of oxalic acid, where they remain until complete decolorization of the permanganate occurs. 5. They are next washed off with sterilized salt solution or water. 6. They are then immersed for two minutes in sublimate solution 1:500. The bacteriological examination of the skin thus treated, even after the mercury has been precipitated, yields almost uniformly negative results, the material for the cultures be- ing taken from underneath and around the nails. This is the procedure now employed in the gynecological and surgical wards of the Johns Hopkins Hospital. Dr. Halsted's method of using subcutaneous sutures, based on the recognition of the impossibility of arriving at a com- plete disinfection of the skin of the patient, diminishes the liability to stitch-hole infection, though even with this method we cannot be sure that the stitch will always be sterile. In conclusion I may be permitted to say that observations of the character reported in this paper are calculated to im- press us with the value of combining sound clinical experi- ence with bacteriological experiment. Our surgical opera- tions are in a sense bacteriological experiments. We may not transfer directly to clinical use the results of bacteriological work, but each serves as a control to the other. We can de- rive from bacteriological observations many useful suggestions and many new points of view. We can guard ourselves often from false deductions from clinical experience by puttingthese AND SKIN DISINFECTION. 15 deductions to the test of bacteriological experiment where this method is applicable. The harmonious working together of bacteriologist, surgeon, and clinician promises results of the greatest practical value.