making the pericardial puncture. The left costoxiphoid angle is preferred by many, who thrust the needle upward and backward from this point. This site is more satisfactory in case of extremely large effusions than under other conditions. The fourth interspace at the left sternal margin or at an inch and a quarter from the margin, as well as the fifth interspace an inch and a half from the edge of the sternum, are respectively advised. Some clinicians in the presence of large effusions are in the habit of aspirating to the left of the left nipple, the effort being made to insert the needle just inside the outer left border of dulness. Personally I am not able to indorse aspiration other than through the small area over the ventricle where the lung does not come in immediate apposition to the chest-wall.

In the event of a purulent effusion surgical measures should be immediately instituted. They consist of free incision, sometimes resection of rib, and the maintenance of continuous drainage, affording a thorough evacuation of thick pus and coagulable flocculi.

ADHERENT PERICARDIUM

There are two forms of adherent pericardium. In one group there is present a simple adhesion of the pericardial and epicardial layers. In these cases the surrounding structures are frequently uninvolved, although there may be a considerable union of the two layers. In another group adhesions are formed between the outer layer of the pericardial sac and the chest-wall, diaphragm, and pleura, by virtue of their immediate contiguity. The parts may be fused closely together and be associated with more or less mediastinal connective-tissue formation.

The symptoms incident to the presence of pericardial adhesions, irrespective of other cardiac lesions, vary according to their extent and situation. In the event of simple adhesions of the two layers of the sac there may be no symptoms whatever, although hypertrophy occasionally may result. Dr. Babcock has called attention to the occasional union of the two surfaces of the pericardium during the time the heart is acutely dilated, either as a result of myocarditis or of valvular disease. Under such circumstances its complete diminution in size is almost impossible, and a disturbance in its function is quite inevitable. When the adhesions are external to the sac and involve neighboring structures, the condition is of much more serious import, owing to the unavoidable restriction in the movements of the heart and consequent embarrassment of its function. Hypertrophy, dilatation, and circulatory stasis are frequent. There are often palpitation, dyspnea, bronchial irritation with weak pulse, and digestive disturbance. The liver is sometimes enormously enlarged from passive congestion. In some cases, however, the liver bcomes much reduced in size through connective-tissue proliferation, and the condition may suggest chronic interstitial hepatitis. The physical signs upon inspection relate to the prominence of the precordium, and a more or less extensive diffusion of the cardiac impulse, with occasional displacement of the apex. The impulse is often undulatory over a wide area, transgressing frequently the limits of the normal precordial region. In some cases there is immobility of the apex impulse, either with a change in the position of the body or during full inspiration. During systole a tugging retraction of the chest-wall is commonly noted in the lower left epigastric region. This may be

followed by a diastolic rebound of the interspaces immediately over the point of apex retraction.

Broadbent's sign consists of a visible systolic retraction of the chestwall, not only in the region of the seventh or eighth ribs in the left parasternal line, but also between the eleventh and twelfth ribs on the left side behind, at the point of attachment of the diaphragm.

Friedreich's sign consists of the diastolic collapse of the cervical veins ascribed to the sudden emptying of these vessels as a result of the expansion of the chest-wall.

A paradoxic pulse, though less common than in acute pericarditis, is sometimes recognized upon palpation, owing to the traction of the cicatricial mediastinal tissue upon the aorta during inspiration.

The percussion signs relate to an increase in the area of cardiac dulness, owing to hypertrophy and dilatation, which may be extreme. Irrespective of the presence of endocardial murmurs, which are occasionally dependent upon the dilatation, the important auscultatory signs are the pleuropericardial friction-sounds. These sounds may be heard both with inspiration and expiration, and, as a rule, are distinctly creaking in character. It is not infrequent to find the sound considerably more intensified during inspiration than expiration, and sometimes vice versa. It often disappears, however, upon holding the breath. It is heard to best advantage over the left border of the heart.

The exact determination of cardiac hypertrophy or dilatation among pulmonary invalids is often a matter of extreme difficulty, even by most skilled examiners. From a comparison of the results of my own clinical findings with the skiagraph I have become convinced, first, that the heart is displaced very much oftener than would be imagined from the literature upon the subject, and, secondly, that the ordinary methods of percussion and auscultation are sometimes quite insufficient to afford an accurate determination of its position and size among this class of patients. This has been discussed more fully in connection with Diagnosis.

CHAPTER LVIII

TUBERCULOSIS OF THE PERITONEUM

TUBERCULOSIS of the peritoneum may exist as one of the local manifestations of acute miliary tuberculosis, or as a distinct peritoneal process of more or less chronic character. When the condition is incident to a general miliary infection, the tubercles are diffused over the parietal and visceral layers of peritoneum, without, as a rule, any active inflammatory change. When tuberculous peritonitis exists as a local condition purely, the inflammatory condition may be accompanied by various pathologic processes. In some cases there is an extensive proliferation of connective tissue with numerous adhesions between intestinal coils and adjacent viscera, with occasionally an implication of the abdominal walls. In others with less tendency to adhesions there is found a pronounced thickening of the peritoneum, omentum, and mesentery incident to their infiltration with degenerative tubercle deposit. This form is sometimes

characterized by the presence of large ulcerative tuberculous masses. Palpable tumors simulating tuberculous growths are produced by a localized matting and drawing of the intestines, which is intensified in some instances by the traction exerted by a shrunken mesentery. In still another group of cases there is a profuse exudative process. This may be unassociated with marked pathologic changes within the peritoneal cavity, or it may attend the proliferative type, with multiple adhesions, or even the caseous and ulcerating forms. The exudation may be serous, seropurulent, or bloody in character, and may be general or sacculated. A simple general ascites of insidious or very acute onset without subjective symptoms may accompany a developing tuberculous deposit unattended by other pathologic change. In some cases the exudative processes are found in association with extensive connectivetissue change. It is manifestly improper to divide all cases of tuberculous peritonitis, aside from the miliary form, into two distinct classes -the proliferative and the exudative groups-as has been attempted.

The etiologic relations of tuberculous peritonitis have been the subject of much clinical study and investigation. The condition is almost always secondary to some other tuberculous focus, the primary source of infection being traced to the lungs in the large majority of cases. At least four-fifths of all instances of tuberculous peritonitis occur in association with easily recognized pulmonary involvement. Cummins has reported a series of cases in which 84 per cent. succeeded pulmonary tuberculosis and 32.6 per cent. intestinal involvement. The same observer has quoted Pribram's report of the result of 165 autopsies upon cases of tuberculous peritonitis, of which 87 were attributed to intestinal tuberculosis, 65 to glandular disease, 8 to tubal and uterine, and 5 to osseous tuberculosis. Douglass has quoted Borschke, who failed to find a primary focus in but 2 cases out of 226 of peritoneal tuberculosis. From a total of 1393 autopsies upon tuberculous subjects, tuberculous peritonitis was found to exist in 226 cases, or 16 per cent. Peritoneal involvement in tuberculous invalids has been reported by other observers to vary from 10 per cent. to 20 per cent.

Tuberculous peritonitis may exist at any time of life, although observed more frequently in young adults. When occurring in children, it is often in association with a general miliary involvement. Frederick C. Shattuck has recently reported some statistical observations upon a series of 98 cases of tuberculous peritonitis treated at the Massachusetts General Hospital during a period of eleven years from 1889 to 1900. The youngest was thirteen months old, while the oldest was sixty-two years. Six occurred in children fron one to five years of age, 7 from five to ten years, 8 from ten to fifteen years, 56 cases, or 57.1 per cent. of the whole number, between the ages of fifteen and thirty years. There were 13 cases between thirty and forty and but 8 cases over forty years of age. In view of the fact that a relatively small number of children. enter the Massachusetts General Hospital, the proportion of cases among children in the series reported by him is perhaps smaller than usual.

Tuberculous peritonitis is usually conceded to exist with greater frequency among females, although some observers maintain that it is more common in the male sex. It would be reasonable to suppose that the statistics should favor the preponderance of the condition among females, on account of their special predisposition to infection through the genital tract. This conclusion seems to be borne out by the obser

vations of surgeons, which, of course, are based upon purely operative

cases.

There are several ways in which the peritoneum is known to become infected. A common method of invasion is from a tuberculous involvement of the wall of the stomach, or of the small intestine, appendix, or colon. The extension of the tuberculous process into the peritoneum from slowly ulcerating deposits upon the intestinal wall in some cases gives rise to purely local changes. In the event of sudden perforation of the tuberculous ulcer a septic general peritonitis rapidly develops through the entrance into the free cavity of the agents of decomposition. The infected area is sometimes sharply circumscribed by the existence of firm adhesions between different coils of intestine. In some instances these adhesions involve the parietal peritoneum.

Another route of infection is that resulting from the caseation and subsequent perforation of mesenteric or retroperitoneal lymph-glands. In such instances the tuberculous process may become diffused throughout the free peritoneal cavity, or it may be localized through the formation of adhesions. An acute septic peritonitis is far less likely to result from this form of infection than from tuberculous intestinal ulcers.

The mesenteric glands have been shown to become infected by the migration of tubercle bacilli through a healthy intestinal wall, an intact mucous membrane constituting no proof that it may not be an atrium of infection. As stated previously with reference to the tonsil, the tissue at the point of invasion may be less favorable for the growth and development of tubercle than more distal parts.

It has not been demonstrated as yet that the peritoneum may become primarily infected by the passage of bacilli through a normal mucous membrane without first producing an involvement of the mesenteric and retroperitoneal glands. The fact that tuberculous peritonitis in exceptional instances has been found to exist in the absence of a discoverable primary focus does not controvert a belief in the secondary nature of the infection.

Another pathway of peritoneal infection is by way of the lymphatics from the pleura through the diaphragm, from the abdominal or pelvic organs or from some distant focus.

The especial frequency of peritoneal tuberculosis in adult females is explained in part by the facilities for direct extension from the female genital organs. The Fallopian tube is frequently a primary focus of infection, from which point the bacilli may gain entrance to the peritoneum, either by direct extension through a free tubal opening or by means of the lymph-channels. It is the consensus of opinion of many observers that the Fallopian tubes are involved in a large proportion of cases of tuberculous peritonitis, which is variously estimated at from 25 to 50 per cent. There seems to be some difference of belief as to whether the involvement of the tube is in general the cause or the result of the peritoneal infection. It is scarcely necessary at this time to review the controversial literature bearing upon the precise direction of the bacillary invasion, whether ascending from the genitals or descending from the peritoneum. It is sufficient to state that the authentic observations of many authorities appear somewhat conflicting and contradictory. Whatever one's theories may be in this matter, it is possible to secure both corroborative and negative testimony as to their correctness. Mayo has called attention to the fact that the tuberculous process in the peritoneum

is especially pronounced in the immediate neighborhood of the primary focus of infection. Certain it is that, irrespective of the source of infection, removal of the tubes is followed in a large number of cases by gratifying improvement. It is also true that in many cases, despite pronounced tuberculous involvement of the tubes, the uterus is found to present an entirely normal appearance. Tuberculous peritonitis has been reported to occur along the route of the male generative tract, although this route of invasion is decidedly less direct than in females. Horowitz describes the pathway from the epididymis through the lymphatics of the spermatic plexus and the ampulla end of the vas deferens. Osler cites seven instances in which the sac alone is involved.

Cruveilhier and Haegler have reported cases of primary hernial tuberculosis. Cotte reports 5 cases of apparently primary tuberculous processes in hernia, together with a summary of 136 recorded cases. It is, of course, easy to understand how hernia may take place in subjects afflicted with tuberculosis of the peritoneum with associated infection of the sac, or of the intestinal coils therein contained, but it is hard to comprehend why there should exist any inherent susceptibility to hernial tuberculosis independent of the peritoneum itself.

Symptoms. The symptoms of tuberculous peritonitis are somewhat variable, according to the type of peritoneal involvement. The condition is of acute onset in but a small proportion of cases. Shattuck reports 29 of acute onset out of a total of 98. Personally, I have seen but 3 cases of acute tuberculous peritonitis, exclusive of the localized infections coëxistent with a tuberculous appendix. In the latter event the onset is often sudden, presenting rapidly fulminating symptoms. Several of my pulmonary invalids have exhibited tuberculous appendicitis with a localized peritoneal invasion of subacute type.

In the majority of instances tuberculous peritonitis is of slow, gradual development. While the course of the disease usually conforms to the chronic type, there may occur for a time acute exacerbations, periods of improvement being followed by recurring relapses. I recall several cases of remarkably slow and insidious onset, which, after the lapse of several years, exhibited periods of severe pain with other acute manifestations which gradually subsided, but were followed by intervals of abdominal discomfort.

Acute cases of tuberculous peritonitis are always characterized by rather extreme pain, and usually by tenderness and tympanites. The pain, however, is not always of an acute character, and in some cases is but little more than a sense of abdominal discomfort, with occasional intercurring colicky attacks. It is increased by intra-abdominal pressure, the tension upon the abdominal wall often being sufficient to produce a voluntary flexion of the thighs upon the abdomen. This position is suggestive of acute general peritonitis.

Nausea and vomiting are quite common in the acute type, but are rarely present in the chronic form, in which the symptoms are rather vague and ill defined. The severity of such acute symptoms as pain, distention, tenderness, nausea, and vomiting, with weak and rapid pulse, is largely dependent upon the extent of peritoneal involvement. In the event of a general peritonitis these symptoms rapidly develop, and are attended by great prostration and followed shortly by a speedy

fatal termination.

In circumscribed tuberculous peritonitis the symptoms are consider