By Oskar Klotz, M. D., C. M.

The discussion to be entered into in this report will be limited to an experience dealing with epidemic influenza as it was met with in the emergency Military Hospital in Pittsburgh. We shall largely confine our attention to the observations which came directly under our supervision, and in as much as this investigation was continued during the epidemic as it swept over this district, the intensive study was limited to a time period of about five weeks. During this period much material was collected, which since then, has taken us a considerable time to analyze. We have thought it more valuable to restrict our discussion to this material in that it illustrates the pathological lesions as they occurred during the acute stage of the disease. We have not entered upon a discussion of the sequelÆ or the chronic lesions which are not uncommonly found following in the wake of an acute epidemic nor do we deal with the lesions arising in cases of sporadic influenza, such as are always with us. As is so well illustrated in the literature, there is probably no disease which has so many late complications and sequelÆ as influenza. The investigations upon the protean lesions have been fully reported in numerous papers during the intervals between epidemics. A comprehensive bibliography upon influenza will be found at the end of the extensive report by Leichtenstern (1905). There is very much less accurate information available upon the actual lesions present during the acute disease when present in epidemic or pandemic form, than upon the many clinical complications in various systems and organs. In fact, our knowledge of the pathology of influenza lies more largely in the field of associated lesions such as the late events in the bronchi, the sinuses of the head, abscesses, meningitis and other conditions, rather to be viewed as complications than as portions of the disease. There are relatively few thorough pathological analyses of the influenza lesions as they are found in the acute epidemic disease.

A fair literature has already appeared upon epidemic influenza from the many countries and regions over which the present pandemic (1918) has swept. These reports by various authors are offered from different viewpoints, some investigators being impressed with certain features which they bring into marked prominence in their reports. It thus happens that up to the present there is a decided lack of uniformity in the opinions expressed upon different phases of the subject. The nature of the pathology of the past epidemic has given rise to many expressions of opinion as well as dogmatic statements, which are found to differ from those of others. It seems to us that this apparent confusion arises partly through the somewhat different characteristics of the disease as it has made its appearance in different centers. We hear it repeatedly stated that the types found in different military camps and urban communities were quite unlike those of other regions. It is evident that such differences in the clinical course actually did exist and that the epidemic though having a common foundation upon which the disease process was built differed in what might be looked upon as symbiotic complications during the early and acute stages. Differences in the nature of the findings in various communities also probably lay in the fact that the bacterial flora associated with the causative agent of influenza was quite different in different regions. We mention this here so that a full appreciation will be obtained for the differences in the pathological characters of the disease as they are found in one region or another. We appreciate, of course, that if the concomitant bacterial flora associated with the underlying cause of influenza, differs in different regions, so, too, will the bodily reactions differ within certain degrees. We are becoming more familiar with different types of bacteria, and the resulting inflammatory reaction which is often unique or at least particular, and that not uncommonly the nature of the inflammatory process suggests the type of bacterium involved. This argument, of course, must not be driven too far, for we well know that the same micro-organisms under different conditions can cause types of inflammatory reactions wholly divergent.

In as much as our observations are confined to a particular group of cases and the study of these was undertaken during the five weeks of the acute epidemic, these results are not to be compared with the collected statistics on influenza as they shall be made over a period beginning with the onset of the epidemic and ending with the last vestiges remaining after months or it may be years of time. Our observations are to be considered only in the light of the events taking place during the height of an epidemic wave. In as much as influenza presents itself during an epidemic in different forms, we shall again mainly limit the report upon our investigations of those cases having respiratory lesions. Our acute observations were made upon the tissues of those who had died of this disease. It is impossible, or nearly so, to fully study the tissues of those with lesser lesions and who recover. Hence, if we divide the influenza cases into those (1) without pulmonary lesions and (2) those with pulmonary lesions, we must state that all of our cases coming to autopsy fall in the second group. It is true that one of these having pulmonary lesions was not brought to his fatal termination by them but by a septicÆmia arising in the middle ear. He had distinct lesions in his lungs. In other words, our autopsy material represents epidemic influenza in which the lung was definitely involved in an inflammatory state. In all but one of these the pulmonary lesion was the cause of death.

No doubt, if opportunity had presented itself to follow a large epidemic through months of its progress, during which late complications in various portions of the body would make their appearance, our analysis would give a different picture and the pulmonary factor for the fatal termination would not be in such prominence.

Of the first group, those cases of epidemic influenza not showing pulmonary lesions, we will have very little to say, in as much as the pathological investigations of them is impossible, or nearly so, during the height of the disease.

Such cases apparently do not die at this period. I am willing to admit that individuals without pulmonary involvement may succumb, but I question whether their death has been due to the result of the influenzal lesions, be it in nose, pharynx, larynx or trachea, or be it in the intestine, but rather that the fatal termination occurred later in the course of this complex disease, when distant vital organs became involved or incapacitated in a toxemia or secondary bacterial invasion. We must clearly distinguish these cases from the clear-cut ones of epidemic influenza, looking upon the new circumstances as complications aside from the original disease. Such, for example, is the case we have mentioned where a fatal streptococcus bacteriÆmia followed in the wake of an otitis media. In our experience we have not had a fatal case of the acute epidemic disease in which the lung was not involved.

In types of epidemic disease such as we have just had, where the epidemic wave has passed over in a period of four or five weeks, there is always much to be regretted which has been left undone. We tried as far as possible to gain all the information available at the time of collecting our materials and of laying aside such of the work which could be accomplished at a subsequent date. The materials were collected from divergent sources in the cadaver, and the more perishable substances were analyzed immediately. During the period of the epidemic 32 autopsies were performed and as much use as possible was made of each for a thorough comprehension of the lesions.

Materials

During the period of our work 639 patients were admitted to the hospital suffering from clinical influenza. The cases varied in type from the very mild to the extremely ill. The majority of the cases were of the type of “three-day fever.” Clinically 81 cases developed pneumonia, and of these, 35 died. It would, of course, be impossible to say how many other individuals had a pulmonary involvement which could not be recognized clinically. In fact, some of the cases which did come to autopsy were only recognized as having a pulmonary involvement when the lungs were examined outside of the body. The physicians freely admitted that the physical signs were quite unusual and unlike those of the ordinary forms of pneumonia. In fact, except for the fact that we were living in the midst of an epidemic of respiratory infections, there was nothing to make the clinician suspect that many of these cases had a pulmonary involvement. Obviously, when the recognized signs of different types of pneumonia made their appearance, the clinician did not fail to make proper interpretation of the lung involvement. This, as we shall discuss later, is an event superadded to a lung condition which pathologically must be recognized as pneumonia (inflammation) and which differs so decidedly from what we know of as croupous or lobar pneumonia, as well as ordinary broncho-pneumonia that it would be incorrect to include them under this heading, although the distribution of the lesion may have lobar, bronchial or lobular characters.

Date 1918		Patients Admitted	Patients Discharged	Cases in Hospital	Deaths
October	5	65	0	65	0
?	6	23	0	88	0
?	7	61	0	149	0
?	8	77	0	225	1
?	9	42	1	266	0
?	10	35	1	300	0
?	11	9	0	307	2
?	12	2	16	290	3
?	13	10	0	298	2
?	14	1	18	278	3
?	15	4	13	266	3
?	16	9	23	248	4
?	17	10	19	235	4
?	18	16	34	217	0
?	19	38	29	225	1
?	20	27	0	252	0
?	21	37	43	245	1
?	22	33	7	270	0
?	23	14	20	263	2
?	24	20	17	266	0
?	25	27	21	272	0
?	26	10	29	250	0
?	27	18	3	265	1
?	28	10	31	243	3
?	29	6	16	231	0
?	30	11	27	215	1
?	31	2	15	202	2
November	1	2	18	185	0
?	2	4	18	170	1
?	3	5	1	174	0
?	4	2	19	156	1
?	5	5	0	161	0
?	6	4	16	149	0
Admissions.		639			35

The individuals admitted to this hospital were obtained from the two military camps at the University of Pittsburgh and the Carnegie School of Technology. All of them were enrolled in the army service and ranged from the ages of 18 to 30. They were vigorous individuals, who had passed their physical examinations for the army. The epidemic made its appearance in these camps on October 2, rapidly ascending from a report of two ill on October 2, four on October 3, eight on October 4, to 65 on October 5. On October 11 there were 307 cases in the hospital.

Of these cases 35 died, the day of death being indicated in the following table.

Day of Disease on Which Death Occurred	Number of Cases
Third	1
Fourth	3
Fifth	4
Sixth	4
Seventh	4
Eighth	5
Ninth	3
Tenth	4
Eleventh	3
Thirteenth	1
Fourteenth	1
Twentieth	1
Twenty-third	1

The time as indicated in the above table has no relation to the length of time that the patients were ill of pneumonia, but refer to the period of illness from the beginning of the influenza. The duration of the pneumonia is indicated in another table.

Of the 35 fatal cases 32 came to autopsy. Facilities were available to do the work very satisfactorily, in that the hospital was well provided with a modern post-mortem room and its accessories. The notes on the autopsies were taken immediately and fully, and the materials for subsequent study were collected in different types of preserving fluid. Portions of tissue were collected from all of the organs for microscopical study, while fluids from the chest, lungs, bronchi and heart were obtained for bacteriological investigations and for some chemical analyses.

Added to the above material we also had the opportunity of reviewing and studying the lesions of 18 autopsies performed by Dr. J. W. McMeans. These cases were very similar to our own series, in that they were cases of epidemic influenza amongst soldiers who were being cared for at the St. Francis Hospital. The disease processes were quite alike in the two series, and the analyses made by Dr. McMeans are comparable in our own and serve as a means of checking our results obtained in another institution. The similarity of the lesions in the lungs and other organs serve to indicate that what is reported in this paper is an index of the nature of the lesions of epidemic influenza as it occurred in the Pittsburgh district. In a few instances the autopsies performed by Dr. McMeans revealed more advanced pulmonary lesions with abscess and gangrene than were noted in the cases autopsied at the Military Hospital. The process, however, in the two series of autopsies was identical.

General External Features

There were no external characteristics of the bodies which were autopsied by us which were constant. Some features were more commonly present than others. Of these the cyanosis of the face, head, neck and shoulders, and in a few instances of the upper extremities, attracted our attention more than any other. This cyanosis was present in over one-half of the number of cases, and it was confined almost always to the upper part of the body. The face, ears and neck were always more affected than other parts. This cyanosis bore no relation to the length of time after death when the body was viewed, as we found that when it was present during life it maintained its prominent appearance for a long time after death.

The cyanosis differed from the bright hue or flush as it is at times observed in ordinary pneumonia, the color in these instances being of a dark purple, or better a purplish blue. The lips and ears showed the most intense color. The cyanosis was not associated with any evidence of oedema. The capillaries of the tissues were filled with blood which was of a very dark character. Cyanosis could also be seen in the finger tips about the nails. This was more marked in the upper extremities than in the lower. The skin of the body rarely showed any cyanosis, these tissues being quite pale, or at times showing a slightly yellowish tinge. In one instance the cyanosis of the head and neck was accompanied by a slight purplish rash upon the upper portion of the chest. This rash was of a petechial kind, there being slight hemorrhage into the tissues. The lesion, however, was not of the blotchy purpuric type which has been observed by others during this and past epidemics (Cole). This single case is the only one where we had evidence of superficial hemorrhages into the skin.

	Cyanosis		No Cyanosis
No.	Degree	Distribution
741	+	Chest and upper extremities	747
743	+ +	Face, neck and ears	748
744	+ + +	Head and neck (upper portion of chest and thighs mottled and purple)	749
745	+ +	Head and neck and upper extremities	751
746	+ +	Ears, neck and shoulders	752
750	+ +	Face, ears and neck	764
756	+ +	Neck, jaw, shoulders and upper extremities	765
757	+	Face, neck, shoulders, arms and chest	778
758	+ + +	Face, ears, neck and upper chest	782
761	+	Face, ears, neck and upper chest	784
762	+	Ears, neck and chest	786
763	+	Head and neck	793
767	+	Face, ears and neck
773	+ +	Neck, ears and cheeks, extending moderately to upper chest. Hemorrhage into conjunctiva
781	+ +	Eyes, lips, ears and neck
783	+ +	Face, lips, neck and fingers
787	+ + +	Ears, neck and shoulders
791	+ +	Ears, neck and upper chest
792	+ +	Ears and back of neck
19		+ Blotchy or slight 6	12 or
or		++ Moderate 10	38.6%
61.4%		+++ Well marked 3
770		Fine petechial rash over upper chest.

Occasionally we met with small hemorrhages lying in the upper layers of the subcutaneous tissue. These lesions were small and could not be seen from the external surface. Nevertheless, some of them seemed to have occurred in direct contact with the deep cutis and surrounded portions of the deep skin appendages. From an examination of our cases there was no reason at the time of autopsy to lay any particular stress upon the occurrence of these hemorrhages. Subsequently, it has come to mind, and since learning of the unusual frequency of boils and deep pustules making their appearance as post-influenzal sequelÆ, that these minute lesions may have a bearing upon the localization of infection in the skin tissues. We must appreciate, of course, that other factors of a constitutional nature probably render the individual more susceptible to the invasion of the staphylococcus, and that such factors are all-important in allowing this organism to gain a foothold. Whether the decreased sugar-tolerance with hyperglycemia, which has been observed in the late stages of influenza, bears a relation to the increased susceptibility, as appears to be the case in diabetes mellitus, is an interesting point for further investigation. Other constitutional states are also undoubtedly involved in the increased susceptibility to the infection which the patient suffers. Elsewhere (Dr. Holman) it is shown that the natural complement content is considerably depressed during the height of the influenza. With such factors present and with the available infecting micro-organisms, it is possible that the minute deep skin hemorrhages bear a relation to the immediate localization of the infection.

In two instances slight hemorrhages were observed into the conjunctival tissues. In each case they were unilateral and occupied the tissues contiguous to the inner canthus. In one case there was well-marked icterus with yellow coloration of the sclerÆ and skin. In this case the icterus was associated with degenerative changes in the liver, there being no recognizable obstruction to the bile passages. The icterus had come on quite acutely and without any special clinical manifestations. In the epidemic of 1890 jaundice was present in a considerable number of cases (Medical Record, 1890, xxxvii, 473). Cole made similar observations in the epidemic of influenza amongst the Canadian soldiers. Œdema of the skin was not met with in any of our cases. This point is worthy of comment, inasmuch as some authors have been impressed with the serious damage taking place in the kidney and the resulting incapacity of these organs. Although, as we shall point out later, the kidney tissues in these cases showed a decided toxic degeneration, there was no evidence that a glomerular damage of serious degree ever occurred. The urinary excretion, as is pointed out in a report by Dr. Zeedick, varies considerably with the intensity of the disease. It is unusual to find derangement of kidney function to a degree to reflect seriously upon the general bodily state. At least this has been our experience in the present epidemic. Even where subsequently we were able to demonstrate a considerable tubular degeneration in the cortex of the kidney the change in the kidney function was not of sufficient magnitude to lead to a water-retention to be recognized in an anasarca. I wish to distinguish clearly at this point the difference in finding an oedema in certain involved tissue structures in various parts of the body and arising through an inflammatory reaction due to the presence of peculiar focal irritation, as compared with the accumulation of fluid in many and irregular situations as it occurs through retention and faulty excretion by the kidneys. Various organs as we have found—as, for instance, the lung, heart and liver—showed a condition of oedema which was not to be reconciled with an inadequate circulation because of a cardiac or renal incompetency. These oedemas, which we will discuss later, are local and are the result of damaging influences inducted in and upon the tissues where they are found.

Muscle

In all of our cases we have been struck with the excellent physique of the individuals succumbing to this epidemic. All were youths in the best of health, of good muscular build and strong bony frame-work. Post-mortem rigidity set in fairly rapidly after death. Where this rigidity had “set” for six or more hours it required much force to change the position of the muscles. The voluntary muscles of the thorax and abdomen were always carefully observed, and in a number of instances the muscles of the thigh were also examined. It was not possible routinely to dissect the muscles of the extremities, so that we are unable to give an accurate account of the occurrence of degenerations in these structures. We have, however, observed the reactions taking place in the pectorals, psoas and muscles of the abdominal parietes. Changes were observed with greatest frequency in the recti of the abdomen. Degeneration occurred in these muscles in 14 instances, while the same tissues suffered rupture, in part or completely with hemorrhage, in six instances. It was not uncommon to find marked degeneration in the lower segment of the rectus muscle on one side, while degeneration and hemorrhage had occurred in its fellow on the opposite side. In four cases rupture of the entire belly of the muscle had taken place, so that a considerable space had occurred between the broken ends and a large clot of blood filled the intervening space. This degeneration, which was seen only in the voluntary muscles, was quite interesting and in its milder degrees was rather difficult to detect. All gradations of loss of muscle color were seen. In some instances the muscle simply seemed to have lost its meaty lustre, while again in the more severe instances the muscle color had changed from the bright red to an insipid yellow or clay color. The most marked degeneration occurred in the midportions, while the ends of the muscle masses at the points of attachment were less involved. Complete rupture of the rectus always occurred in the lowermost segment, a short distance above the insertion into the pubic bone. At times the distribution of the degeneration within the muscle was quite patchy, and irregular islands of yellow about 2 cm. in diameter were splashed through the muscle masses, which in themselves were paler than normal. Where the muscle degeneration was advanced the tissue was soft and at times even buttery. It resembled the character of the degeneration observed in typhoid fever, although I have no recollection amongst many enteric cases of having seen the degeneration of the muscle occur so acutely. Recklinghausen claimed that these hemorrhages were most unusual in influenza. This is contrary to our findings.

Degenerations of a similar kind as those of the abdominal recti were found in both pectorals. In the chest region, however, the degeneration was less frequent and less severe. We observed it only twice, and in neither instance had the degeneration led to a rupture and hemorrhage of the muscle bundles. Kuskow observed a single case of degeneration and hemorrhage of the pectoral muscles. In the psoas muscle we observed degeneration on two occasions, in one of which the lesion was associated with a partial separation of the muscle fibers and hemorrhages into its substance. In one case clinically, but not coming to autopsy, a lesion, which from its character we presume to have been a degeneration, occurred in the sterno-mastoid, being accompanied by hemorrhage and the development of a firm clot the size of a hazel nut. In the subsequent history of this case the lesion passed through an aseptic process of organization with contracture so that the patient has recently been developing a “wryneck.” Kohts in 1890 reported the finding of muscle degeneration and abscesses in the arm. The condition arose as a late complication of influenza.

From our experience at the autopsy table in observing the relative frequency with which muscle degeneration occurs in the severe cases of epidemic influenza, we feel convinced that numerous cases which recover pass undiagnosed of this condition. Furthermore we have evidence, as illustrated in a case observed by Dr. McMeans, wherein a lesion which occurred in the gluteal muscles was followed by a localizing infection at this site that these muscle degenerations and hemorrhages may have serious consequences. There are a number of instances in which post-influenzal complications of the nature of deep-seated abscesses of the extremities, thorax, and abdomen may have their explanation for the localization in a primary muscle damage accompanied by hemorrhage and followed by an infection of variable type. Cole also comments upon the development of abscess in the deep muscles where degeneration had taken place. In illustrating some of our findings to Dr. J. Anderson he immediately recognized such a condition in the pectoral muscles of a patient in which he was unable to arrive at a conclusion of the pathological events which had taken place. It is one of the noteworthy features in this disease that the voluntary muscles of certain regions are apt to suffer severe damage, while the heart and the various unstriped muscular tissues are little if at all affected by a similar process. It would be interesting to know whether the lack of response and the delayed functional recovery on the part of the muscles of the extremities in so many patients who have suffered influenza is the result of the damaging influence of a peculiar intoxication present in this disease. One of the features in influenza is the prostration of the patient, and with it there is definite muscular weakness. We have been prone to lay the responsibility of this state entirely at the door of the nervous tissues. Here, however, we are able to offer evidence that quite aside from the lesions arising in the nervous tissue, there is definite muscle damage which, as we shall again discuss when describing the microscopic features, incapacitates even to the point of complete destruction the muscle elements in various fields of the body. Before, however, being able to state that the muscular weakness of the extremities is the result of such damage by toxins it is necessary to obtain more definite information regarding the frequency with which these degenerations occur in the limbs. In our own material we are unable to discuss the matter with adequate figures. We are, however, impressed with the changes observed in the muscles which were available to us. Naturally, too, a certain number of muscle degenerations have escaped our detection because of our unfamiliarity with the mildest grades. In fact, we have already discovered in our microscopic studies that certain cases, which in the macroscopic had escaped us, showed well-marked lesions under the microscope.

Abdominal Recti		Pectoral	Psoas
Toxic Degeneration	Hemorrhage Into Rectus	Toxic Degeneration	Toxic Degeneration
745 on 10th day	745 both on 10th day	756 on 8th day	756 on 8th day
749 on 4th	752 both on 13th	770 on 11th	792 on 6th
752 on 13th	756 both on 8th
756 on 8th	764 both on 9th
757 on 6th	765 both on 9th
762 on 10th	778 both on 23d
763 on 11th
764 on 9th	Rupture of
765 on 9th	Rectus
767 on 10th
770 on 11th	745 right on 10th day
778 on 23d	756 both on 8th day
783 on 8th	778 right on 23d day
791 on 6th

We have convinced ourselves that the marked hemorrhage taking place in the muscle tissue follows upon a primary degeneration of this tissue and its spontaneous rupture. The amount of hemorrhage is in proportion to the degeneration and fracture of the muscle elements. The hemorrhage does not precede the muscular change, nor does it have any antecedent relation to the actual tearing of the muscle fibers.

A much better appreciation of the muscle degeneration was obtained in the microscopic studies of these tissues. The various gradations of tissue change could be followed, which was not possible in the naked-eye examinations. Some points respecting this degeneration were quite noteworthy. Firstly, the process of degeneration in its early stages and advancing through the acute destructive periods was not accompanied by any inflammatory reaction. Evidence of inflammatory exudate was obtained only when the degeneration had proceeded to a degree permitting of rupture with hemorrhage, or in the late stages when the areas of marked muscle dissolution were undergoing repair. We have no evidence to indicate that bacteria were present during the beginning of the degenerative process. Bacteria could not be demonstrated in section. The appearance of the tissue suggested a purely toxic process which was selective in its action, picking out voluntary striped muscle tissue and attacking certain muscle groups in preference to others. It was also interesting to observe in the early stages of the degeneration that individual fibers lying amidst healthy and unchanged muscle elements would show degeneration in many of its stages. This appearance was often unique, particularly when in the early stages of the process the involved fiber would still retain its normal position and shape though markedly altered in its staining and chemical qualities.

The degeneration as observed in these cases showed many of the characters like that of waxy degeneration seen in typhoid fever. Similar appearances to these have also been described in connection with the toxic degenerations which occur in the vicinity of infections by the gas bacillus. In fact, all the stages observed in the one can be seen in the other. They differ, however, only in the degree to which final destruction takes place and in the speed with which the degeneration is accomplished. The character of the degeneration is well studied in sections stained with hematoxylin and eosin, eosin-methylene blue, and best of all in the phosphotungstic acid hematoxylin. By the latter method one is able to follow clearly the grade of degeneration as it effects the muscle striations. On the other hand, the peculiar waxy appearance of the early degenerating fibers is best seen in sections stained with eosin or fuchsin, where the striated muscle fibers are found to be changed to a more intensely staining red body of homogeneous character and devoid of all evidence of their original internal architecture. These bland waxy fibers were often of the size and shape like the normal. On the other hand, the fibers are also not uncommonly swollen, stretching the sarcolemma to almost the bursting point. Following this primary bland degeneration the fiber takes on irregular shapes, becoming constricted and collapsed at irregular intervals, so that islands of the waxy contents lie within the sarcolemma, being separated from each other by constricted areas in which the original myoplasm has undergone decomposition and sometimes complete absorption. This irregular destruction of the muscle contents often has a granular stage in which the original muscle substance has become disintegrated. The sarcolemma follows the condition within it, stretching when the fiber is swollen and shrinking, or even becoming collapsed when the inner substance is becoming liquified and absorbed. The sarcolemma does not suffer the degenerative changes of the inner fiber, nor can one observe nuclear changes in this sheath which are significant.

When first studying this process of degeneration it appeared to us that the earliest change was a loss of the transverse striations and the subsequent disappearance of the longitudinal fibrillÆ. We have subsequently found that this is incorrect and that the changes observed in the markings of the fibers were not constant. At times the muscle substance would progress through stages of degeneration up to the point of disintegration and dissolution while the transverse striÆ were still discernible in the altered fiber. The one constant change that we have observed in the degenerating fibers was the early loss of staining qualities as obtained by the phosphotungstic acid hematoxylin. In such preparations the earliest effect of the intoxication upon the muscle fiber was a change in reaction to this stain. Sometimes within a given fiber small irregular and poorly staining blotches could be observed, while the remaining portion of the fiber was normal in its appearance. Later these poorly staining areas became larger, occupying the entire width of the fiber and being distributed at irregular intervals in its length. Finally the characteristic staining quality was entirely lost, although in the poorly colored cell transverse striations were still discernible and a true waxy stage had not yet taken place.

At times the waxy degeneration advanced into the stage of disintegration by an irregular destruction within the fiber. When this occurred the fragments of waxy substance took on curious coiled and grotesque shapes, while a granular destruction was taking place in their periphery. Neither inflammation, oedema nor a vascular reaction could be determined in these tissues of mild or severe change. The reaction as is indicated in the table occurred quite acutely and was not accompanied by fatty products commonly seen in the slower forms of degeneration.

Gradually the debris of the degenerated fibers is absorbed and the sarcolemma shrinks and collapses upon itself. During this stage a reaction occurs in the sarcolemma with nuclear proliferation. At times the last vestiges of the muscle fiber are seen to be surrounded by a crown of nuclei and cells reminding one of the appearance of the degenerating nerve cells in the Gasserian ganglion in hydrophobia. The involved area becomes active in appearance, showing proliferation of fibroblasts and the appearance of occasional lymphocytes and plasma cells. Scar tissue continues to develop in proportion to the amount of damage done. In areas where hemorrhage had taken place the amount of scar tissue is exaggerated, owing to a process of organization which is taking place quite apart from the muscle degeneration. Thus not a few scars scattered through the voluntary striped muscles are the final outcome of this toxic degeneration occurring in epidemic influenza. Some of these lesions may account for the indefinite pains and symptoms of which the patient complains for so many months after his acute illness. I refer particularly to lesions occurring in the psoas and muscles of the back as possible explanations for the partial invaliding of some individuals.

In a certain number of cases of acute influenza the patients complain of severe abdominal pain, in the absence of any localizing symptoms or evidence of intestinal derangement. Such was the case with a number of the above cases coming to autopsy, and the sole evidence we could offer was muscle degeneration with or without massive hemorrhage. The abdominal pains complained of were more of the nature of dull aches with occasional exacerbations and shooting or lancinating “stitches.” Rarely was the patient able to define the position of the pain, not being able to state whether it was within the abdomen or in the parietes. Most frequently they claimed it was internal. We have on no occasion demonstrated an intra-abdominal lesion which could account for such pains. None of our cases was of the type of “intestinal influenza.” We are, therefore, led to the conclusion that the muscle degenerations of the various degrees, from the slight with few muscle elements involved to the severe with rupture and hemorrhage, account for a proportion of the clinical symptoms of (muscle) pains and aches as well as weakness. We cannot claim that coughing was a necessary factor in inducing rupture of the abdominal recti. In some of the cases with rupture severe coughing had not been observed during the illness.

Upper Respiratory Tract

The pathological changes found in the nose, pharynx and larynx were of relatively slight importance and most variable in their severity and incidence. The majority of individuals had few clinical manifestations of disease in these parts. Some, however, complained of dryness of the pharynx with slight feeling of fullness. An examination of these parts revealed some congestion, varying from a red injected mucosa to a bluish cyanosis. In the nose the reaction was rarely as acute as is seen in infectious coryza, but even where relatively little change was to be seen in the tissues hemorrhage from the erectile tissue was not uncommon during the acute stages. No particular lesion was to be found associated with nose bleed. There was an unusual absence of excessive secretion from nose and pharynx in the majority of cases. One was also struck with the infrequency with which the larynx was involved. A certain number of individuals complained of hoarseness, and in them injection of the vocal cords with some swelling was found. In many others, however, even where an intense infectious process was present in the lower respiratory tract the larynx was almost without change. It was from the level below the larynx that the acute reaction in the respiratory system was found.

In all of our cases the trachea showed definite inflammatory reaction. Of the 32 cases there were 26 having an acute tracheitis, 5 with an acute mucopurulent inflammation and 1 with a reaction in the subacute stage. In the majority of the cases with acute tracheitis there was a thin layer of exudate lying upon the mucosal surface. At times the trachea was filled with a frothy serous fluid, the greater part of which had its origin in the lung. Nevertheless, as we shall point out later, we did obtain microscopical evidence indicating that during the early acute stage of the tracheitis a considerable serous exudate escapes from its mucosa. This serous inflammatory reaction is an important one for all of the mucosal structures upon which the virus of influenza obtains a footing. This we have found true for the trachea, bronchi and alveoli of the lungs. In some cases the exudate was grey and lay in close contact with the injected tissues. At first sight this grey exudate suggested necrosis, but it was readily wiped from the underlying structure. Some leucocytes and cell debris with many bacteria made up the content of this grey exudate.

The macroscopic appearance of the trachea was that of an intensely injected structure which had largely lost its normal lustre. The naked eye could distinguish that anatomical change had occurred in the surface tissue of the trachea and that there was unusual evidence of intensely injected vessels lying in the submucosa. In only one instance was there an appearance of a true necrotic membrane lying upon the surface of this intensely inflamed layer. This apparent membrane was found to consist of a wide patch of desquamated epithelial cells which was lying as a delicate necrotic plate upon the surface. This thin layer was devoid of a meshwork of fibrin threads as usually accompanies a true false membrane of other sources.

The early intense inflammatory reaction of the surface membrane of the trachea was characteristic, and in our experience was never exceeded in intensity by other infections. A desquamation of the lining membrane was also a common finding. Naturally this intense reaction so commonly found in the trachea extended without interruption into the main bronchi and their divisions. The finding of this continuous surface inflammation is good evidence of the mode of spread of the infectious process along these membranes, beginning in the upper portions and by direct continuity involving more and more of the respiratory tubes toward the lung.

The varying grades in the intensity of the inflammatory reaction upon the inner surface of the trachea was well illustrated in the microscopic sections. Even with the different degrees of the reaction there was a fairly constant character to the inflammation. In this way the response was found to differ from that commonly observed in ordinary infections of the respiratory tract. The first striking feature is the marked response of the vascular channels, both blood and lymphatic. The vessels lying in the submucosa were found intensely engorged so that their walls were stretched to the point of bursting. In fact, not a few vessels were seen whose walls, probably under the stress of intoxication and dilatation, had given way leading to a flooding of the neighboring tissue with their contents. Where such vessels lay close underneath the surface the hemorrhage escaped into the lumen of the trachea. Accompanying this early vascular response there was found a marked serous exudate leading to a stretching of the submucosal tissues by distention of the interstitial spaces. This reaction resembled an acute inflammatory oedema and occupied the area between the mucosa and the inner border of the cartilage rings. Beyond this region no response was found. Thus in the earliest stages, and where the mucosa was still intact, the main reaction was of the nature of an intense serous inflammation with congestion of the blood vessels and frequent interstitial hemorrhages.

Shortly following the development of the serous exudate in the submucosal tissues, the epithelial lining is found to suffer from the reaction. The serous exudate does not remain confined to the interstitial tissues, but is poured out through the mucosa into the trachea. It would appear that the amount of this clear exudate may become greater than can be dealt with by the mucosa, with the result that an accumulation of this serous fluid takes place between this epithelial layer and its basement membrane. We have repeatedly seen considerable stretches of the mucosa lifted from the basement membrane and shed in large plaques into the lumen. These mucosal cells at the time of their desquamation retain fairly well their morphological characters, and do not show evidence of necrosis prior to their removal. Disintegration of these cells naturally occurs while lying in the secretion of the trachea, and a variable cellular mass in stages of disintegration may often be found both in smears and sections. When the epithelial cells are lifted in wide plates, a type of bleb develops which is easily broken and then disintegrates.

The desquamation of the lining membrane is a fairly constant occurrence in the cases coming to autopsy. In the majority of those which we have examined the greater portion of the trachea was completely denuded, save for small islands lying in the recesses near the mouths of the mucous ducts. In one case this lesion was accompanied by a process of ulceration, due in all probability to the invasion by other micro-organisms. The denuded tracheal surface usually shows a further inflammatory reaction in which a cellular exudate then makes its appearance. This reaction is mainly one in which lymphocytes and plasma cells infiltrate the spaces previously occupied by the serous fluid. The reaction is limited to the submucosa and does not extend into the tissues beyond the cartilages. We have found only occasional polymorphonuclear leucocytes lying close below the surface. During this period, however, varying grades of degeneration may occupy the upper layers. The basement membrane particularly seems to suffer by losing its characteristic outline and staining qualities. This membrane becomes swollen, softened and indefinite. At times a homogeneous precipitate occurs along its free surface giving rise to an appearance resembling a false membrane. This deposit is, however, distinctively different from the diphtheritic membrane of other infections. It is interesting, however, that where such deposits and degeneration occur in the basement membrane more or less degeneration and necrosis also occur in the connective tissues immediately neighboring to it. These tissues show a peculiar granular destruction and alter their staining qualities. Moreover, and what is more important, under these conditions the dilated blood vessels are found to suffer from the injuries taking place in their neighborhood. We have repeatedly found partially or completely thrombosed capillaries, arterioles and venules in these surface layers. These thromboses took place while the vessel was in its distended state and thus produced a mold of the dilated vessel. This observation is of importance in indicating the severity of the effect of the virus and toxin upon the tissues of the trachea, and it is also of importance to appreciate that this damaging influence is very different from that which we encounter in pneumococcus infections, and we shall point out in our discussion on lung a reaction very similar to that which takes place very superficially in the trachea may also occur in the alveolar walls of the lung.

Having referred to the intensity of the responses of the blood vascular system, we must also indicate the part played by the lymphatics. Simultaneously with the reactions taking place about the blood vessels of the trachea we observed similar responses in the lymphatic channels. At first these dilated structures contained only fluid. Later the migration of the lymphocytes took place along these routes, and rarely micro-organisms could be demonstrated either free or within an occasional leucocyte. The sharp response of the lymphatics during the serous inflammation is noteworthy, inasmuch as we have found that the lymph glands lying about the respiratory tubes and lungs were early in their response to the irritating virus.

Bacteria were demonstrated in the secretions lying upon the surface of the trachea. In those specimens in which the mucous membrane was still intact we attempted to demonstrate the clustering of the micro-organisms about the ciliated cells as was described by Mallory in whooping cough. Although the organisms, and particularly small Gram negative bacilli, could be demonstrated lying about these cells no characteristic arrangement was found. Furthermore where the mucosa was still attached to its basement membrane we were never able to demonstrate organisms below the surface of the epithelial layer. In several cases where the mucosa was lifted in bleb-like structures a number of organisms were detected below the epithelial layer and in contact with the basement membrane of the submucosa. We have rarely demonstrated bacteria in the interstitial spaces of the submucosa, even where large numbers of organisms were lying upon the inner denuded surface.

The distinction which was made by the gross examination of the trachea between the acute tracheitis with serous exudate, subacute tracheitis and mucopurulent tracheitis was not so readily distinguished in the microscopic sections. In the gross the character of the exudate lying upon the surface was the main guide suggesting the nature and intensity of the inflammatory reaction. In the microscopic sections this exudate was largely wanting, or was not sufficiently characteristic to confirm the gross findings. On the other hand, differences in the nature of the injury were to be found mainly in the reaction of the submucosa. As we have indicated above, the early inflammatory reaction of the trachea is mainly evident in an intense congestion accompanied by an inflammatory oedema of the submucosal tissues, hemorrhage sometimes accompanying this response. In the later stages of the reaction a cellular deposit takes the place of the inflammatory oedema and usually consists of lymphocytes and plasma cells. It is only in those cases where the intensity of the irritant continues to act over a longer period of time that a superficial necrosis with leucocytic infiltration makes its appearance. The epithelial layer of the trachea is desquamated early in the acute reaction, and hence a denudation of the surface is to be found in all stages of the acute lesion. The mucous glands have not been found to show any particular involvement in the inflammatory process, and in the majority of instances they were found to have escaped entirely the damaging effect of the virus. Their response in an over-secretion of mucus may be the outcome of a stimulation by toxins or soluble irritants; but on the other hand, may also probably be a reflex response to the injury of the mucosal surface, which being bared of its covering is highly sensitive. The increased discharge of mucus from the deep glands may well be a protective response to such injury.

Bronchi

The lesions in the bronchi were in every way comparable to those in the trachea. The main bronchial tubes differ in no material way from the structure of the trachea, and the extension of the inflammatory process from above downwards leads to a reaction in their walls similar to what has been above described. As we follow the subdivisions of the bronchi we gradually lose some of the characteristics contained in the larger tubes. The mucous glands gradually become fewer and eventually disappear. The cartilage rings become smaller and no longer completely encircle the bronchus, and with the further diminution in the size of these structures disappear entirely. A relatively greater amount of muscle tissues takes the place of the cartilage rings. This change in the anatomy of these structures has a certain influence in modifying the character and distribution of the inflammation.

Thus whereas we have indicated that the inflammation of the trachea and of the large bronchi is of a peculiar kind and remains confined to the tissue lying inwardly from the cartilage rings, we found that where these structures give place to a loose muscle tissue with a more extensive lymphatic drainage the zone of inflammation is not so limited, but proceeds outwardly into the neighboring tissues. We often use the terms bronchus and bronchioles very freely without clearly distinguishing any real difference. In a study of the inflammatory reactions of the respiratory tubes in epidemic influenza (as well as in other infections) it is best to accept the anatomical definition that the bronchioles not only represent the minute tubules passing to the alveoli, but also those small air passages which devoid of cartilage, mucous glands and heavy connective tissue stroma are in close relation to the parenchymatous tissues of the lung. These soft muscular tubes possess blood and lymphatic vessels which freely communicate with the blood vessels of the lung alveoli. It is in association with these distant tubes that concomitant inflammatory reactions are found in the alveoli and in the bronchial tubes.

Desquamation of the epithelial lining is to be found in every size of bronchial tube where the infection has caused an acute inflammatory reaction. Throughout the pulmonary tissues where the lung is found in some stage of influenzal pneumonia the bronchial tubes, both large and small, are either entirely denuded of the mucosa or show only remnants attached to irregular areas. In the smaller passages dense clusters of desquamated cells are sometimes found within the lumen and indicate the accumulation of a desquamated epithelium obtained from portions of the tubular system in deeper portions of the lung. In the early stages, this desquamation is accompanied by a serous exudate and a certain amount of hemorrhage. Later we find masses of leucocytes which fill up the tube, and though appearing to arise from these structures have in fact largely come from the lung alveoli. Like the larger bronchial tubes the distant ramifications show relatively little cellular reaction in their walls in the early period. It is only when the neighboring lung tissues are extensively implicated in a purulent inflammation that we find a similar exudate occupying the tissues of the bronchioles. Polymorphonuclear leucocytes are equally distributed through the region of the basement membrane, submucosa, muscular coat and outer connective tissue layer. Some grades of degeneration may occupy the inner surface wherein the basement membrane first shows a homogeneous swelling and later a granular degeneration. In a few instances where the small bronchioles have communicated with regions with abscess formation an ulcerating surface occupied the inner boundary.

The evidence in the smaller bronchial tubes, both those with cartilage and those without, that an inflammatory reaction of some degree may occupy the muscular coat is of importance. We have found reactions of inflammation in the muscular coat varying from a mild oedema and cellular exudate to an intense polymorphonuclear leucocyte involvement. In the latter the muscle fibers showed evidence of degenerative change and suggested an acute weakening of this layer. We lay particular importance upon this finding as indicating a causative factor in the development of acute bronchiectasis as was met with in one of our cases. In this particular instance the bronchi passing to the lower lobes of each lung were unusually dilated and could be followed, in the gross, to their distant extremities. The dilatation was more or less uniform and no large pouches or cavities had developed. A mucopurulent exudate was found occupying these dilated tubes. Others have likewise observed the development of acute bronchiectasis under these conditions. Goodpasture and Burnett found that as early as the second to the fourth day one of the striking appearances was the gaping dilated condition of the infundibula, and the tendency to dilatation of the air passages was manifested in a bronchiectasis in 4 out of 30 cases. Boggs as well as Lord have reported upon chronic bronchiectasis associated with the B. influenzÆ and there appeared to be evidence that a certain percentage of cases recovering from influenza permanently develop irregular dilatations of the bronchial tubes.

The recognition of inflamed bronchi or bronchioles was never difficult. In the gross the presence of the abnormal exudate and the intense injection of the mucosal surfaces always attracted attention to the inflammatory state. Furthermore where the mucosa had been desquamated the surface of these tubes was found to be quite granular if closely observed. With moderate magnification by means of a hand lens the granular appearance was shown to be due to the engorged vessels. Much easier, of course, was the recognition of the inflammatory reaction by the microscope. The importance, however, of the bronchitis and bronchiolitis lay in the amount of involvement which had occurred in the neighboring tissues. As we, however, indicated elsewhere, we do not doubt that many of the cases of three-day fever have a state of tracheitis and bronchitis equal to that which we have observed in many of our cases. Whether the inflammatory reaction progressed beyond the firmer bronchial tubes to the softer and more vascular structures would be difficult to say where our evidence rests upon the clinical findings alone. It is, however, probable that a certain number of the severe and sharp attacks of influenza not only cause a tracheitis and bronchitis of the larger tubes, but also extend more deeply into the smaller ramifications tending to simulate the reactions which we have above described. When we ask ourselves, however, how distantly must the infection invade the smaller bronchial tubes before involving the parenchymatous tissues of the lung we are at a loss to enunciate a general rule. It is more than probable that there are modifying influences which determine whether the bronchitis with a certain amount of its bronchiolitis will progress to a true pneumonia or will remain localized to these tubular systems. I can well appreciate that in the event that a bronchitis has an inflammatory reaction accompanied by much serous exudate there is great danger of flooding the neighboring alveoli with this inflammatory fluid and of carrying the large numbers of the micro-organisms within the tubes to the air sacs of the lung. Under these conditions the virus has an unusual ability to develop the disease from one localized in the air passages to that of a true pneumonia. It is probable that the peculiar early acute reaction which is present in the air passages in epidemic influenza is responsible for the extensive involvement of the lung in the severe and dangerous form of inflammation.

S—Serous. H—Hemorrhagic. P—Purulent. B.P.—Broncho-pneumonia. S.F.—Serofibrinous. F.—Fibrinous. F.P.—Fibrinopurulent.

It was very evident that the smaller bronchi and bronchioles were much more readily involved in a severe inflammatory reaction than the larger tubes. A purulent inflammation was not uncommonly found in the bronchioles of the lung when a pneumonic state with leucocytic infiltration was present. Even where such purulent infiltration of the walls of the bronchioles was readily demonstrable the trachea and main bronchi were devoid of this intense reaction. These purulent inflammations were not uniformly distributed in the bronchioles of the lung, but only occurred in those regions where the parenchymatous tissues were in themselves involved in a purulent reaction. It was difficult to find the evidence whether the purulent bronchitis preceded or followed the presence of a purulent pneumonia. The intimacy of the lung tissues with those of the small bronchioles makes it impossible for one or other of these structures to escape when one of them is implicated in a purulent reaction. It is equally important to appreciate that to a considerable extent the lung tissue surrounding the small bronchioles becomes involved by a direct radial extension through the walls of the thin respiratory tubes. Such extension laterally is assisted by the free lymphatic communication lying about the bronchioles and stretching into the lung parenchyma. Purulent processes of the small air tubes always showed a similar reaction in the interstitial tissues of the neighboring air sacs.

Our material did not permit of following the bronchial reactions to their conclusion. In some instances we have found that where abscesses developed within the lung the contiguous bronchi and bronchioles either became eroded or suffered intense suppurative inflammatory lesions on their inner surface. The manner in which repair of the more common inflammatory processes of the bronchi is accomplished could not be demonstrated in the cases dying during the acute stage. In one case an organizing bronchitis was associated with an organizing lobular pneumonia. In this instance the connective tissues were proliferating freely from the inner wall of the bronchi, there being no evidence of a basement membrane at the point where the connective tissue was growing. The development of the connective tissue appeared to be spontaneous and was not taking place within an unresolved fibrinous exudate. In as much as the fibrosing process was largely scattered through all of the lobes, the numerical involvement of the respiratory tubes was quite great. In this instance the amount of obstruction which was imposed upon the respiratory tissues by the fibrosing pneumonia and bronchitis was sufficient to cause considerable distress and dyspnoea during the last few days of the patient’s life. The amount of dyspnoea was out of proportion to the clinical manifestations of pulmonary involvement, and from a clinical point of view it was difficult to arrive at a conclusion of the nature of the lung lesion.

Undoubtedly during the subsidence of the inflammatory process within the bronchi the gradual restitution of the tissues with little or no fibrosis is accompanied by a reproduction of the lining membrane arising from the epithelial remnants in the small mucous crypts. In a few cases lately coming to autopsy where the patients had suffered an influenza five or six weeks previously, the mucosa of the trachea and bronchi had assumed its normal appearance and was fully clothed by a normal epithelial covering.

Lung—Early Stage

We have just discussed the importance of the inflammation of the trachea and bronchi in the cases of influenza. It is our belief that every case of influenza has some tracheitis, and a great many have both tracheitis and bronchitis. This is true in the absence of localizing signs and symptoms, as was evident even in these cases in which the simple influenza passed into its more severe type with its pulmonary lesions. In many of these instances clinical evidences of an inflammatory reaction in the respiratory tubes were wanting, while the reactions observed at autopsy were often astounding.

Just as we feel that simple influenza and inflammation of the respiratory tubes go hand in hand, or better that these respiratory localizations are the all-important ones in every case of simple influenza, so, too, we are of the belief that the pulmonary lesions bear the same relation to all cases of severe and fatal epidemic influenza. We hold that no case comes to his death through acute epidemic influenza without having a lesion in the lung. The pulmonary condition, therefore, is of first importance and its analysis is imperative for a proper understanding of this disease. There has been divided opinion as to the part played by the pulmonary lesion in epidemic influenza, some holding that it is to be looked upon as a part of the disease and others that it must be viewed as a complicating lesion. Complications of various kinds are very common, and there are a number of conditions arising in the lung (abscess, gangrene, necrosis) which must be viewed as complications. There is, however, a type of pneumonia, and here I use the term in its broad sense, which is not in truth a complication but merely a wider extent of involvement of the respiratory tract by the same virus which is always present to cause lesions in the respiratory tubes. The reaction within the lungs is distinctive and differs from the pneumonias which are met with under other conditions and with various bacterial agencies. Nor are our findings in this matter unique for this epidemic. They have been described and discussed in the past. True it is that, like in the epidemic which has just passed us, the incidence of clinical and pathological pneumonia varied quite widely in different communities, so, too, the reports of past epidemics do not give a uniform description of a pulmonary lesion. Where, however, the analysis has been made during the four weeks’ period of the acute epidemic and where the descriptions have been recorded by painstaking observers, the similarity with our present findings is very striking. I would refer in particular to one report made in 1893 in Petrograd by Kuskow. His report deals with 40 carefully studied cases in which records both macroscopic and microscopic were accurately made.

One of the great difficulties in placing an accurate interpretation upon the pulmonary findings lies in the fact that true pneumonia as seen in epidemic influenza in man has not been reproduced in animals. Furthermore, as the majority of the fatal human cases of epidemic influenza with their associated pneumonias present a mixed infection of the lung tissues, it is difficult, if not impossible, to indicate the lesions which have resulted through the activity of one of these as against those induced by the other bacteria present. In our own carefully studied cases wherein bacteriological cultures were taken from every lung there was not a single instance in which the influenza bacillus was present in pure culture. This is more fully commented upon in the studies by Dr. Holman, but the point we wish to make here is the difficulty in arriving at a conclusion in our material as to the actual effects induced by any one type of organism. As it is fully discussed by Dr. Holman we are convinced of the importance of the influenza bacillus in this epidemic. We also appreciate that pneumonia lesions in animals have been induced by a variety of materials gained from influenza patients, but yet in view of the abnormal manner of producing such lesions these are hardly comparable to those in man. We may well expect severe oedema, inflammation and hemorrhage, if in guinea pigs, rabbits and monkeys we introduce by intra-tracheal insufflation large quantities of fluid suspensions of bacteria. And thus we find positive results obtained by the use of a filtrable virus, streptococci, influenza bacilli and other organisms. The lung is a sensitive tissue which quite readily responds to a variety of irritants. In many respects some of these lesions simulate those in influenza, but still we are far from the conclusion that the disease, influenza, with all its manifestations has been actually reproduced.

The pathology of the pulmonary lesions in acute epidemic influenza is so distinctive that except for the late purulent stage which may resemble types of reinfected and unresolved pneumonia the condition cannot be confused with the stages of frank lobar pneumonia. We appreciate that this is a very positive statement, and that opposition will be taken by those who resting their opinion upon individual factors may claim that a clear distinction from other forms of pneumonia is not available. We, however, base our opinion not upon a single feature, but upon the combined pathological complex observed in many individual cases. These features are mainly those seen in the type of the lesion, the character of the distribution, extent of involvement and the multiple stages so commonly present at one time in different portions of the lung. The type lesion that has become so well known in pneumococcus lobar pneumonia has its distinctive stages which for teaching purposes are divided into the stage of (1) congestion, (2) red hepatization, (3) gray hepatization and (4) resolution. In dealing with lobar pneumonia from the standpoint of illustrating these stages the majority of teachers annually confess their inability to present for the student’s study the stage of congestion. The student is impressed that the congestive stage of lobar pneumonia is very transient and rapidly passes into the stage of red hepatization. Patients do not die with pneumococcus pneumonia in the stage of congestion. And this is also largely true of the stage of red hepatization, which is but rarely seen at the autopsy table. This community (Pittsburgh) gives its large quota to the mortality statistics of pneumococcus pneumonia, but it is most unusual to meet with a specimen of red hepatization except for the borders of the advancing gray area. And, furthermore, red hepatization even when found in the unusual cases shows remarkably little of this character when seen under the microscope. True it is that a certain number of red blood cells will be found in the alveoli and a certain degree of congestion will occupy the alveolar walls, but its extent is far less than what we may have hoped to demonstrate to others. So that broadly speaking the intensely congested lung with or without red hepatization is unusual in our frank lobar pneumonia. This was quite the reverse in our cases of acute epidemic influenza-pneumonia. Furthermore lobar pneumonia in the great majority of instances illustrates a distribution distinctive for the name. Massive lobar, or pneumococcus pneumonia is found to occupy one or more lobes or parts of lobes. The involved lobe is fairly uniform in the stage of the inflammatory process. If it is in the early gray stage, this will be seen with equal intensity in the different areas of the lobe. Patches of pneumonia in different stages within the same lobe are not to be found, while this finding is not uncommon in the pneumonias of acute epidemic influenza. And lastly, the frequency with which an inflammatory oedema occupied the lungs in the cases of influenza was in quite striking contrast with the dry fibrinous lesion of common pneumonia. This wet state of the lung was but a stage in the inflammatory process varying in its extent in the different periods, but nevertheless inducing a character in the early pulmonary lesions which was quite foreign to our usual finding. This wet state also assisted in modifying the subsequent picture so that when the lung assumed its gray appearance it was rather of a slimy character than of the firm dry nature. In this late gray stage the slimy lung somewhat resembled the appearance of unresolved pneumonia where this condition had been brought about by a new infection upon the original cause of the pneumonia.

It is incorrect in influenza pneumonia to speak of the lesions as lobar pneumonia or broncho-pneumonia if by these terms we have in mind the pathological characters observed in the pneumococcic pneumonia with its lobar or bronchial distribution. Influenza-pneumonia appeared with both lobar and lobular characteristics. Nearly every case had both types of lesions present, but the nature of the inflammatory process is so decidedly different from that of the ordinary endemic pneumonia that a confusion in the interpretation is likely to arise and in fact has already raised a considerable polemic. Influenza-pneumonia is commonly lobar, lobular or bronchial in distribution. It is, however, not of the characters that are associated with the lesions designated under these terms. When, therefore, we here use the word “lobar” we mean lobar in distribution but not lobar in type. As will be seen from our table, it was usual to have multiple lobes involved. But the lesions, not only in the different lobes varied in their character and distribution, but even within the same lobe a variety of types was present.

Day of Pneumonia on Which Death Occurred	No. of Cases
Second	2
Third	4
Fourth	7
Fifth	6
Sixth	7
Seventh	3
Eighth	1
Tenth	1
Twentieth	1

To a certain degree we were able to analyze the types of the lesions as they occurred in the different stages and progress of the pulmonary inflammation. Briefly, these were as follows: the earliest stage of congestion following rapidly upon the infection from the bronchi was followed by (1) inflammatory oedema, (2) hemorrhage, (3) cellular exudate (a. mononuclear cells, b. leucocytes, c. interstitial infiltration) and (4) resolution or organization, abscess, infarct and gangrene. The majority of our cases died during the stages of congestion, hemorrhage or early purulent infiltration. In the early stages the amount of fibrin was small or entirely absent, later, with the appearance of leucocytes, some fibrin was present.

For the estimation of the time elapsing between the onset of the pneumonia and death we are dependent upon the clinician. This is often quite difficult to do, in as much as with a primary respiratory disease, such as epidemic influenza represents, it is very difficult to determine the time when there is a transition from the inflammatory process of the upper respiratory tubes to that of the pulmonary tissue. In many of the cases where from the onset there was intense prostration and every evidence of marked intoxication the clinical manifestations of localized processes taking place in the respiratory system were very much in the background and often of insidious progress. In four of our cases it appeared as if the pulmonary manifestations had made their appearance with the first sudden and severe onset of the influenza. On the other hand, also, the clinical signs and symptoms of lung involvement were different from those of frank lobar pneumonia. We would, from our experience at the autopsy table, say that where in the cases of epidemic pneumonia there are present the signs of pulmonary consolidation like those of true lobar pneumonia, that there has been an antecedent period of a pulmonary lesion which passed unrecognized by the clinician. To more clearly state the case, whereas in lobar pneumonia the stage of congestion preceding the stage of red hepatization gives rise to no signs whereby the clinician can indicate the time of its onset or determine the time when it has passed into the succeeding stage, and moreover, the stage of congestion is of short duration to be measured in a period of a few hours, this stage in epidemic influenza though equally indefinite in its clinical manifestations is much prolonged, lasting not only a period of hours but even a period of several days. It is this pulmonary state which is difficult or even impossible to recognize in the living. All gradations of it occur and the clinician can only broadly suggest from all the evidence at hand, the period when inflammation with definite exudate began in the lung. In as much as the total length of illness of a number of cases was only three, four and five days, whereas there was nothing at the onset to suggest pulmonary involvement, we can estimate approximately, at least, the duration of the lung condition. This makes it possible to give a relative estimate of the character of the lesions present at different periods of time. The outstanding finding, as we will discuss again, was that a distinct and peculiar pulmonary reaction was primarily imposed upon the lung, which made its appearance at periods different from those of frank lobar pneumonia.

We were repeatedly surprised at finding death to have occurred during the stage of acute congestion with some hemorrhage and inflammatory oedema of lung and in the absence of any sign of grey hepatization or purulent infiltration. In many of these cases the involved areas of lung though heavy and oedematous, were still partly air-containing and the amount of lung involvement was insufficient, on the basis of mechanical interference, in accounting for the severity of the clinical symptoms and the fatal outcome. This must have impressed everyone dealing with the autopsies during the acute epidemic. It immediately suggests that in some cases at least the pulmonary lesion, in as far as incapacitating the external respiratory system, was not the sole or even the important cause of death, but that a condition of intoxication, borne out by the evidence of damage in muscles, blood and kidney is a large factor of danger in this disease.

We shall briefly describe the important pulmonary findings as we have met with them in the successive stages of influenza-pneumonia. This, we hope, will make clear the interpretation of the pathology of the lung lesion of the epidemic as it came under our observation.

The earliest pulmonary lesion which we encountered was one of congestion, inflammatory oedema and hemorrhage. These three conditions were usually present at the same time and were found in the height of intensity in all of the cases dying within the first four days of illness. During this early period these manifestations of inflammation were not accompanied by definite red or grey hepatization as might ordinarily be expected. The lesions varied greatly in their intensity, the oedema always being very prominent, while the hemorrhage varied from a diffuse infiltration of the involved lobe or added to this, was localized in massive collections four or five cm. in diameter and commonly occupying the central portions of the lobes. We have seen several hemorrhages lying in close proximity to each other with their borders coalescing and leading to a larger central involvement. In the regions where the hemorrhage and inflammatory oedema were diffuse, air was still present within the lung tissue, sometimes to an extent permitting the lung tissue to float on water but more often in quantity sufficient only to suspend the tissue at various depths. On pressure the fine air bubbles were recognized amidst the blood-stained fluid. Acute compensatory emphysema often occupied the anterior borders of the lobes or formed interstitial blebs beneath the pleura. The quantity of fluid, inflammatory oedema and hemorrhage, contained within these bulky lobes was often very surprising. A lobe when compressed would leak fluid with the ease that it could be obtained from a sponge. Out of the lower lobe on one occasion we pressed 700 c.c. of limpid blood-stained exudate. The acute emphysema which may make its appearance suddenly, is

Degree of Involvement	+	++	+++	Total
Left upper lobe	10	6	7	23
Left lower lobe	4	10	17	31
Right upper lobe	12	8	9	29
Right middle lobe	10	9	4	23
Right lower lobe	5	9	18	3
All lobes were simultaneously involved in some grade of pneumonia in 18 cases—56 per cent.

at times quite remarkable. It may appear very early in disease. We have not met with a single case where the emphysema of the lung led to a rupture of the air sacs and an interstitial infiltration of air through lung, mediastinum, neck and subcutaneous tissues. Some very remarkable cases are reported by different authors where this emphysema was of astounding grade leading to a crepitating infiltration throughout the mediastinum, neck and the subcutaneous tissues over the thorax and abdomen as low as the pubis. The milder grade of emphysema consisted mainly of an abnormal expansion of the air sacs which were not infiltrated by exudate and which probably had some effect in preventing the diffusion of the inflammatory fluid from entering certain regions. These emphysematous areas could be readily recognized by the naked eye along the anterior borders of the lung as well as between the involved pneumonic patches within the lung.

These lungs, involved in this early serous and hemorrhagic exudate varied considerably in their appearance according to the regional and quantitative involvement. As is seen from Table viii, the lower lobes were more commonly occupied by massive exudate than the upper, and the involvement of multiple lobes was the usual. Still more remarkable is the fact that all lobes were simultaneously involved in some grade of reaction (pneumonia) in 56 per cent. of cases. In complicated influenza-pneumonia Goodpasture and Burnett found the inflammatory reaction in both lungs and involving to a greater or less degree the lobes on each side. Most commonly this involvement consisted of a lobar distribution in one or two lobes with a lobular or patchy disposition of exudate in one or more of the remaining lobes. Where the distribution was lobar the involved lobe was distended to its fullest and the pleura tightly stretched over the lung tissue which, heavy with fluid, was not solid but flabby. The lung could be moulded under the finger and could be compressed into various shapes. At first sight this flabby, heavy lung tissue suggested the appearance of the waterlogged lung which one encounters in renal disease or failing circulation. A closer analysis, and particularly when the lung was sliced, showed an entirely different character.

Where the inflammatory oedema was accompanied by much focal hemorrhage the distribution was nodular and suggested the appearance of the hemorrhagic lung of plague pneumonia. It was this appearance which led to the suggestion that the pandemic was not one of influenza but possibly of an infection related to the eastern plague. The nodular masses of hemorrhage at times occupied areas varying from the size of a walnut to that of a golf ball and were localized amidst a relatively mildly involved lung tissue making a sharp contrast between the involved and relatively normal tissue. With the removal of the lung from the body and the partial collapse of the aerated tissues these nodules became still more prominent. The greater the amount of hemorrhage within these areas the more solid became the occupied tissue. Such sporadic distribution of hemorrhagic lesions occurred in the two most intense and rapidly fatal cases. Both of these individuals died within 48 hours of the time of onset of the lung conditions. In these two cases we do not believe that the pulmonary lesions had been prolonged over a time even as long as 36 hours but with the difficulty of estimating the onset of the lung involvement we are giving a liberal estimate of this time.

Besides meeting with the stages of congestion, oedema and hemorrhage during the earliest days of the pulmonary lesions we have found that they are to be encountered virtually through all the stages of the fatal cases either as remnants of the original reactions which had not been entirely obliterated by the succeeding purulent process or as was so commonly found, new reactions occurred in other regions of the lung so that, in the same individual, inflammatory reactions of different stages of development could be defined. I do not recollect a single autopsy of a case dying during the acute period which did not show evidence of some areas in the stages of this early acute reaction. Naturally where resolution is well advanced within the lung all trace of inflammatory exudate of various kinds is removed and where such individuals with their resolved pneumonia are brought to death through succeeding complications the above finding will not be borne out. We limit, however, our statement to the findings in the acute deaths.

We have previously intimated that the oedema present in the early stages of the reaction is to be looked upon as an inflammatory oedema or better as a true serous exudate, and must not be confused with the transudation of fluids in non-inflammatory conditions. We have on several occasions collected the fluid expressed from the soggy lungs and have made some determinations of their chemical qualities. The difficulty immediately arises in separating the materials arising from cellular degeneration from the natural constituents of the serous exudate. We were unable to obtain specimens in which laked blood was not present, so that even though the cellular constituents and fibrin were removed, decomposition products could not be separated. The analyses, however, gave a differentiation from the transudate seen in renal and cardiac conditions.

During the period of the accumulation of this inflammatory fluid the clinician could often recognize a profuse watery exudate within the lung or even observed an abundant serous discharge arising in bronchi and trachea. At times the quantity of expectoration was great. Frothy serous fluid accumulated in the air passages and would periodically be expectorated. At other times the hemorrhage was of quite serious extent and the patient would suddenly bring up several mouthfuls or more of bright blood. This pulmonary hemorrhage was without manifestations different from the acute illness with cyanosis of other individuals. The two most acute cases, which we have referred to above, were of this kind, both of them having marked hÆmoptysis with the loss of upwards of a pint of blood at a time.

The early pulmonary lesion which we have described, we have called acute serous pneumonia and acute hemorrhagic pneumonia (or we might speak of it as an acute sero-hemorrhagic pneumonia) and is one which is distinctive for epidemic influenza. The cut surface of a lobe involved in this reaction is wet, glassy, meaty and oozes much blood-stained fluid. It contains no visible fibrin and presents no characters of a “cellular consolidation.” As a serous inflammation of the lung it is unique. The further remarkable character to the pulmonary lesion is that in advancing through the other stages, it never passes through a stage of “red hepatization.” Here again we have a distinctive difference from the pneumococcus-pneumonia. From what we have previously said about the nature of this early acute inflammation of the lung in this disease it is apparent that red hepatization has no place in its process. The stage of red hepatization is attained only when the inflammatory reaction is accompanied by certain constituents in the exudate, which upon coagulation (separation out of the fibrin) renders the lobe dry and solid, while there is a sufficient abundance of red blood cells and congestion to maintain a dark red color. The hepatized lung on section is dry, more or less granular, containing fibrin, red cells and leucocytes within the alveoli. Extensive oedema is unusual except in the cases of hypostatic pneumonia, which in well marked cases bears some resemblance to the gross appearance of the early influenza pneumonia. We have not encountered a single case of the red meaty lung of influenza which showed evidence of true red hepatization in the gross.

The microscopical examination of the lung tissue confirmed the observations which were made in the gross. In the early stages of congestion the reaction was much more extensive than what could be spoken of as a broncho-pneumonia. The capillary dilatation in the alveolar walls occupied diffuse areas varying from multiple lobules and areas several cm. in size to the common diffuse congestion of an entire lobe. Capillaries were distended to their full capacity and often this engorgement was associated with the leakage of blood or a serous fluid. Not uncommonly a clear serous fluid was exuded into the interstitial tissues of the alveolar wall and collected within the air sacs. The high albuminous content of this fluid was seen in the homogeneous coagulation which occurred when the tissues were placed in fixatives. The microscopical sections of such parts demonstrated the coagulum occupying the alveoli as a clear homogeneous substance containing relatively few cells and looking not unlike the colloid deposit of the thyroid. The alveolar walls, themselves, were infiltrated with fluid so that the distended tissues and vessels made these structures thick and bulky. In our own observations we were impressed by the differences of the early inflammatory reaction from those ordinarily seen in pneumonia. Amongst these differences was the quantity of fluid extruded into the lung with a relative absence of fibrin. In some instances fibrin was completely wanting, although small quantities could be demonstrated in isolated areas. This observation upon the quantity of fibrin can be made only during the early stage of the disease in as much as after secondary infection of various kinds has become implanted the presence of fibrin has become a variable quantity often exceeding that seen in the early stages. This is one of the points upon which the older authors have laid stress in differentiating influenza pneumonia from others. In this we fully concur. Whether this lack of fibrin in the inflammatory exudate is a characteristic to be associated with the infection by the B. influenzÆ alone is hard to say, but in as much as it was such a prominent finding we are led to lay some stress upon it. It is, of course, to be realized, as with all other micro-organisms that under certain conditions fibrin will form an important part of the exudate even when the B. influenzÆ is present. This is true in the inflammatory reactions of the meninges present in infections due to this bacillus. Under the conditions of epidemic influenza where the lung lesion is the prominent and unique reaction this micro-organism fails by itself to bring out this quality in the exudate.

Not uncommonly this stage of inflammatory oedema was accompanied by various grades of hemorrhage, varying from the presence of small aggregations of red cells to a complete flooding of the lung tissue making it look not unlike a red infarct of lung, save that the alveolar walls still showed an active circulation and living cells. It was remarkable that even though there was such an intense reaction taking place in the lung tissue there was little or no evidence of a cellular exudate during this stage of the process. Where much blood was extruded into the alveoli occasional fibrin threads were found in the coagulum. In these early cases the bronchicles and small bronchi were found to contain an exudate similar to that in the alveoli. Not uncommonly the vessels from which the red blood was escaping, could be demonstrated in sections. The appearance of the vascular wall suggested that a definite opening had occurred in the side of the capillary from which the blood escaped. We were not able to demonstrate a fatty or other type of degeneration in the cells of the capillary walls. It is probable that the process of injury was much too acute to permit of the demonstration of the products of degeneration within the surviving cells.

The hemorrhagic lesions which had existed for a longer period of time gradually showed a varying infiltration by wandering cells. The earliest cells not belonging to those of the hemorrhage or oedema appearing within the alveoli were mononuclear elements partly arising from the alveolar walls and partly coming from the circulation. Numerous mononuclear cells of epithelial type desquamating from the inner surface of the alveoli accumulated in the oedematous fluid and the hemorrhage within a short time after their occurrence. These cells either appeared in clusters or as single elements. Accompanying this were also large mononuclear cells loaded with different quantities of pigment which had apparently escaped from the lymphatic channels within the alveolar walls. These latter cells belong to the wandering endothelial type which are active in phagocytosis for foreign material and which assist so largely in inducing the deposit of carbon in the lungs and lymph glands. A third mononuclear cell appearing early in the reaction was the lymphocyte. The numbers and extent of distribution of this cell were not constant. We have seen it in some of the reactions where very few leucocytes were to be seen, and where it constituted the main infiltrating cell of the alveolar wall or the air sacs. We have previously mentioned its presence in the inflammatory reactions of the bronchi. Here we find it in the early response within the lung tissue and appearing amidst a reaction which is intensely acute. It is not long after the finding of these various cell elements that the polymorphonuclear leucocyte wanders in large droves to numerically overshadow the mononuclear cells. Nevertheless, the three types above mentioned can be recognized in the exudate through the succeeding stages of reactions in the lung. The large macrophage shows its phagocytic properties in taking up numerous red blood cells, lymphocytes and occasional leucocytes.

It is not difficult to demonstrate that the inflammatory reaction within the bronchi and bronchioles precedes the responses within the alveoli. Quite often one may find an acute bronchiolitis with desquamation of the lining epithelium and the early serous exudate lying amidst the lung parenchyma unaffected by any irritant and reaction. There is every evidence that the bacteria reach the lung tissue by extending along the walls of the respiratory tubes and eventually reaching the air sacs either in the distant extremities of the bronchioles or when they have arrived at the thin-walled structures extend through them into the neighboring air sacs.

It is during this early period that we are able to observe the characteristics of the initial inflammatory exudate as we have described it above. The serous exudate and the infiltration by mononuclear cells appear early while the absence of fibrin also attracts attention. In place of fibrin there appeared in a certain number of cases a peculiar material of a hyaline nature which becomes plastered against the borders of the air sacs forming a fairly thick laminated structure and within which thread-formation is not to be seen. Occasionally a few cells lie within this hyaline substance. Some have referred to this as a type of fibrin. We have found, however, that it does not give the staining reactions for fibrin and does not appear to be of the same composition. These masses are tightly welded to the alveolar walls and the borders are often indistinguishable. In part this material appeared to be made up of necrotic cells of the septum which previously had suffered oedema and circulatory interference. We have found in a number of cases hyaline thromboses of the fine capillaries with more or less necrosis of the alveolar septum. At times the septum was entirely destroyed so that a thick hyaline mass alone separated neighboring air sacs. This hyaline necrosis resembles in part the superficial necrosis which was observed along the borders of the denuded bronchi. There is, however, more than necrosis of cells constituting this deposit for the bulk of material eventually deposited is much greater than could arise from tissue cells alone. These hyaline masses have never been found to lie upon the alveolar wall with an intact lining, but it is always accompanied by a loss of the lining cells and more or less destruction of the wall itself. As to the nature of the hyaline deposit which is laid down in lamellae we do not know. Fibrin threads occasionally appear to arise from these hyaline deposits and extend amidst the exudate in the air sac. One cannot assume, however, that the fibrin and the hyaline material have any relation to each other as their chemical characteristics (and mode of deposition) appear to be quite different. It has been suggested by some that this hyaline material represents an imperfectly formed fibrin which has formed a jelly-like clot, not having the property of developing the usual threads.

It is of importance to appreciate that the deposition of these hyaline structures indicates a severe injury of the alveolar walls not commonly observed in ordinary pneumonias.

In different areas of the same lung these constituents of the early exudate may be observed in all proportions of admixture. Each one of the elements of the exudate may largely overshadow the others and prominently modify the appearance of the lesions. Broadly speaking, however, the inflammatory oedema and hemorrhage occupying the greatest part of the exudate in the lungs and the absence of marked leucocytic response as well as the absence of the characteristic fibrinous meshwork in the alveoli give to the early influenza-pneumonia a character different from those which we ordinarily see.

It is during this early phase of the reaction that the influenza bacilli can be shown within the lung structures. The distribution of bacteria is not uniform. Clusters of these minute bacilli are found in the alveoli at irregular intervals, many of the air sacs containing much exudate being quite free from organisms. When present the bacteria appeared in tightly aggregated schools lying free amongst cells of the exudate, but also certain numbers being incorporated within the large mononuclear cells. In some regions organisms of the type of the influenza bacilli were alone seen, while elsewhere again, and particularly where the exudate was assuming purulent characters other bacteria of the nature of streptococci, staphylococci and micrococcus catarrhalis, were also found.

Lung—Secondary Stage

Following upon the primary reaction in the lung as above described, a secondary reaction makes its appearance at variable periods. This reaction is one in which the inflammatory exudate resembles more closely but is not identical with the responses which are observed in ordinary lobar, lobular and pneumococcus-pneumonia. Whereas in the earlier period, the reaction is largely one of a serous and hemorrhagic exudate accompanied by peculiar hyaline deposits along the inner borders of the alveoli, later there is seen a change in the quality of the exudate with the accumulation of more cellular elements and some fibrin. The naked eye appearance of the involved tissue changes considerably. The lung tissue loses in weight but becomes more solid. The lung contains less fluid and the cut surfaces are drier and the color of the reaction changes from the dark congested appearance to one showing all varieties of red and gray. This change from the flabby and soggy pneumonia to the more definite type of consolidation occurs in the regions which have been previously involved and is not to be found in the lung areas which have escaped the early reaction. The gray consolidation appears to be either a stage of the influenza-pneumonia or is a new reaction superadded to those pulmonary lesions induced by the primary infection.

It is sometimes difficult to recognize the beginning of this pneumonic stage inasmuch as the gray color does not make its appearance even with the presence of fairly large quantities of cellular exudate. The amount of hemorrhage that originally lay in the affected areas for a long time overshadows the presence of the color of the cellular exudate. This is also true of the characters that may be impressed by the presence of fibrin. Small quantities of fibrin scattered through the congested and oedematous lung are not readily recognized and the beginning of this secondary reaction is also easily overlooked if one relies upon evidence of consolidation. More or less solid exudate may occupy a flabby lung without permitting one to appreciate its presence in the gross specimen. When, however, the deposit is of sufficient quantity to change the color of the involved lobe and to alter its consistency, one has little difficulty in recognizing the changes now taking place. The earliest development of this change in the inflammatory reaction was on the fourth day. In the majority of instances the gray color and the consolidation made its appearance about the sixth day. We have, however, on several occasions observed hemorrhagic lesions as late as the seventh and eighth day, at which time it was impossible to recognize a gray hue to the exudate or the character of granular consolidation to the involved lung.

The reaction naturally suggests the stage of gray hepatization as we so well appreciate it in ordinary pneumococcus-pneumonia and from the standpoint of its color and the greater solidification of the lung tissue we might speak of it as such. Here, however, it must be clearly distinguished from the gray hepatization of ordinary pneumonia. This secondary lesion of influenza-pneumonia has but little in common other than its color and the development of a consolidation with true lobar pneumonia. It is never as clear cut as we see it in the latter and the degree of the “gray hepatization” is not uniformly distributed through the involved lobe. One portion of the lobe will show a diffuse gray hue while in other parts more decided lobular or patchy areas are picked out in the advanced reaction. There is not the uniformity of lobar involvement nor is the distribution as regular as one obtains it in broncho-pneumonia. Furthermore, the character of the consolidation differs very decidedly in showing such a variety of hues in reds and grays and the cut surface is not the picture of the dry granular consolidation of our endemic disease. The gray areas are in all states of wetness and ooze a slimy fluid on the cut surface. In the later stages this exudate is most profuse resembling a sticky pus. In its appearance we were reminded of the character seen in unresolved pneumonia as well as in the pneumonias produced by the pneumococcus mucosus, and the B. mucosus capsulatus. We would, therefore, avoid the use of the term gray hepatization and in place of it, as the evidence with the microscope confirms, use the term purulent pneumonia.

There are three other characters which differentiate this gray stage from those of ordinary pneumonias—(1) the irregular distribution, (2) the friability of the involved tissue and (3) the interstitial reaction. We have never observed such an irregularity in the distribution of a gray stage of pneumonia as we have seen it develop in acute influenza-pneumonia. All types of involvement of the lobes are found in different cases and even sometimes in the same case. The least frequent type has been the broncho-pneumonia in its true form. Broncho-pneumonia as we see it in children and the cases following measles is usually fairly uniformly seeded through several lobes and the size of the individual patches is about that of a split pea. The small bronchus can be recognized about the center of the involvement. In those instances one has studded through the lung tissue numerous small swollen areas which are granular, dry and gray. Differing from this the patchy distribution of the gray stage of influenza-pneumonia had no regularity either in the size of the areas nor the distribution. A lobe may show one or more patches. The patches may be distributed toward one portion of the lobe more than another. Furthermore the areas do not always encircle the small bronchi but involve the terminal portion so that an entire lobule is more commonly affected. The lobular type rather than the peribronchial type is most commonly seen and it is often remarkable how sharply the gray lobule is demarcated from the surrounding congested lung tissue. On several occasions we observed a single lobule in the gray stage while the remaining portion of the lobe was in the serous and hemorrhagic condition. However, multiple lobules are commonly seen closely associated in the advancing inflammatory process. Such lobules show peculiar geographical patches or leaflet-like configuration. Varying with the number of lobules involved the extent of the gray change in the lobes assumed more or less a lobar distribution. There was no uniform position to this pneumonic state sometimes appearing in the peripheral tissues of the lung, at other times lying centrally with less involved or less advanced inflammatory reactions surrounding it. Nevertheless, the gray stage made its appearance more rapidly in the lower lobe than the upper and it was not uncommon to find this condition appearing quite early in the upper posterior portion of the lower lobes. This latter position is the one which is recognized during life by the clinician as one of the earliest localizations of the demonstrable pneumonia. It is reported by many that the first physical signs of consolidation are to be obtained close to the lower angles of the scapulae.

There is no doubt that the character of the pneumonic process in the epidemic influenza was not the same in all localities. There have been not a few who have reported a large proportion of their pulmonary lesions as a definite broncho-pneumonia with an interstitial purulent involvement. The prominent reaction was a small circumscribed yellow focus about the bronchioles from which a bead of pus could be expressed. These pea-sized foci were scattered through several or all lobes. It is this type of reaction which appears to develop by a direct extension through the bronchial walls and to remain quite localized in the alveoli about these tubes. This reaction seems to be purulent from its very beginning and does not pass through the stages as we have described them above. There is more or less fibrin present in the exudate, but usually not in the quantity observed in lobar pneumonia. These lesions closely resemble those observed in the post-measles pneumonia, and it is claimed are the result of the same agent; the hemolytic streptococcus. In only one case did we observe a lesion of this kind. The small areas of broncho-pneumonia were confined to the left lower lobe and in the lower portion of the upper lobe. Each area was about the size of a split pea, was quite yellow and in fairly sharp contrast to the background of an acute sero-hemorrhagic pneumonia. The subsequent history of these interstitial purulent broncho-pneumonias is like that in measles, where the tendency toward an organizing pneumonia has been shown. The importance of the hemolytic streptococcus in inducing purulent interstitial lesions of the lung (and also of other organs) cannot be over-impressed. It is not so much the type of the reaction during its acute stage which attracts our attention, but the manner of the healing process. It is more than probable that the organizing pneumonias of influenza, not only of this distinct bronchial type, but also the lobular, confluent and lobar variety have had an associated streptococcus infection. The more intimate discussion of this type of pneumonia has been given by MacCallum.

Our autopsy experience has led us to believe that the definite clinical signs of pneumonia are associated with the development of this gray consolidation of the lung. The lung tissue develops characters which permit the physical signs to be recognized. The tissue is more solid and more readily transmits the bronchial sounds. This is not true of the earlier stages where the inflammatory process is contained within a lung tissue which still is partially crepitant and when the so-called consolidation is due to an inflammatory oedema and not to the more solid fibrinous and cellular exudate. With the protean distribution of the gray lesion one does not wonder at the clinical difficulties in mapping out or even finding the consolidated tissues.

As soon as the lobes show this gray character and with the progressive development of an acute interstitial purulent pneumonia, the lung tissue becomes friable. All gradations of flabbiness may still be obtained and in the early stages while the cellular exudate is accumulating to change the color of the lung, little variation from the tough character of the pulmonary tissues can be recognized. When, however, a true gray character is assumed by a portion of the lobe, the tissue becomes so soft that it is handled with difficulty without rupture. The thumb can be pressed into the gray mass and pus will well up around the invading phalanx. The consistency in the late stages reminds one of the pulpy tissues in acute splenitis. In cutting such lobes it is almost impossible to obtain slices of the tissues, their own weight often breaking such a segment. When allowed to rest on the table for a few moments, the cut surface becomes coated with a dirty yellow slime representing pus and products of disintegration arising from the lung. The stroma and alveolar tissues are themselves involved in the inflammatory process and many of them have suffered complete or partial destruction so that they offer but little resistance to pressure and serve as a poor supporting stroma to the pulmonary tissues. The reaction which has taken place within the lung producing both the gray color and the destruction of the tissues is, indeed, an active suppurative one. One would not be surprised to obtain not only a purulent lesion wherein the cellular exudate occupies the air sacs and their walls but also a further stage leading to a destruction of the tissues to the extent that abscess cavities are produced. These we have met with in several instances, some of them being small while others were several centimeters in diameter. An abscess of larger extent and having a destructive process which involved the surrounding tissues so that one would speak of it as a process of gangrene, was observed by Dr. McMeans in one of his cases. A lobar distribution of the purulent lesion takes place where multiple involved lobules have fused in their periphery or where a suppurative flooding of the tissues in this violent late reaction has taken place.

The question at once comes to mind whether this gray stage is but the late event of what we have previously spoken of as influenza-pneumonia or whether this condition is superadded to what may begin as an influenza-pneumonia but end in a pulmonary inflammation with a mixed infection. Dr. Holman was not able to demonstrate a sufficient difference in the bacteriology of the lobes in the gray stages from those in the early acute stage to be able to say that the flora changes at a certain time during the progress of the disease in the individuals. It is possible, and there is some evidence in support of this, that the earlier stages of the pneumonic process represent the reaction to the influenza bacillus and that during this period the response is fairly uniform and similar owing to the fact that this infection has but a short incubation period and a high pathogenicity. In such an event the particular micro-organism may bring about a peculiar response of its own before the other organisms with which it is associated have the opportunity of producing damage. Subsequently, however, these secondary organisms impose their peculiar reactions upon an altered lung, thus inducing an inflammatory lesion which differs from the preceding reaction and also differs from the reaction usually induced by those organisms upon relatively healthy tissues. It is difficult to account for the very irregular distribution of the gray lesions by an explanation concerning the influenza bacillus alone, or by the characters peculiar to the secondary infection. There is an entire want of character to these gray lesions which makes them differ from other types of pneumonia known to us.

It is well to lay particular stress upon this peculiarity in the distribution and extent of the lesions within the lobes; and it is also important to appreciate the difference in the appearance of these gray areas from those of true lobar or broncho-pneumonia.

Finally there is another point in which this stage of the pneumonic process differs from that of pneumococcus lobar pneumonia. In frank lobar pneumonia the reactions taking place in the involved portion of the lung are fairly uniform in all its parts. The stage of red hepatization occupies about that amount of lung which subsequently shows itself in the state of gray hepatization. In other words, all of those areas which appear gray are preceded by this peculiar red consolidation, and all of the area occupied by the red hepatization will pass through the phases of gray hepatization before entering upon the final stage of resolution.

In influenza-pneumonia, on the other hand, the events taking place in a given lobe are not uniform and various stages and grades of the inflammatory reaction may be recognized at the same time, some appearing red, some congested, some flooded with blood in hemorrhage and others showing the purulent infiltration by the appearance of gray patches upon the background of red. Not only do the various reactions within the same lobe fail to show similar grades of intensity and similar stages or time of involvement, but we find that all of the red and hemorrhagic areas are not destined to pass through the gray stages. At times it is true an entire lung will enter into the purulent phase and if this becomes extreme abscess and gangrene are almost certain to develop. But often the purulent infiltration occupies only a few or scattered lobules and resolution may take place in a lung where the greater part of the lobes is occupied by the inflammatory oedema and hemorrhage and has never become truly consolidated by cellular and fibrinous exudate. This feature that the involved lung tissues need not pass through the sequence of events which is usually observed in frank lobar pneumonia is so distinctive that it differentiates the character of the inflammatory reaction very clearly. It may be that this is an indication of the unequal distribution of the micro-organism and that the first infection presumably by the bacillus influenzÆ has been much more diffuse and of wider extent than the secondary invading bacteria which being distributed through the bronchial tree are more or less localized to those lobules most severely involved. It is impossible to claim for influenza-pneumonia as clear and sharp-cut stages as we obtain them in the pneumococcus lobar pneumonia.

During the period of the intense purulent reaction in certain portions of the lung, the intrinsic structures within the area also partake in the damage and response. The suppurative infiltration not only occupies the alveolar walls but also extends through the tissues of the bronchioles, the arteries and the veins. The polymorphonuclear leucocytes seem to migrate into all of the parenchyma indicating some damage by bacterial invasion. On more than one occasion have we observed partial or incomplete thrombosis of arterioles and capillaries whose walls showed an acute suppurative reaction. Some of these thromboses are of importance, being associated with the interference with a blood supply not compensated by adequate anastomosis. Necrosis and small areas of gangrene and abscess are to be found in the region of the circulatory disturbances. It is also during this period of the disease when the bronchi and their ramifications contain pus or muco-pus, that the exudate from the alveoli readily finds its way into the air passages and becoming mixed with the mucus from these tracts forms a tenacious discharge.

The presence of large amounts of exudate within the bronchi brought these structures into unusual prominence. This was particularly true in the purulent stage of the reaction when beads of sticky pus would well up from the cut bronchioles. We were tempted on a number of occasions to speak of this in terms of bronchiectasis but with the intense inflammatory reaction occupying the bronchial wall and modifying its contour on this account we avoided this diagnosis. In one instance, however, the lesion was unmistakable. This was a case of purulent pneumonia (764) dying on the ninth day of the disease. The distribution of his pulmonary lesions was distinctly lobular, apparently following the course of the bronchial distribution. The bronchi were followed longitudinally and irregular pouchings of the lumen were very apparent. The bronchi had suffered marked inflammatory reaction which had also infiltrated the muscular tissues of the tubes. Goodpasture and Burnett report finding two cases of acute bronchiectasis associated with abscess and ulceration of the bronchi. In our case the bronchiectasis was found bilateral but was more marked in the lower lobes than the upper.

The lymphatic channels within the lung tissue are found active in establishing an internal drainage to the neighboring thoracic glands. The lymph vessels were often found filled with leucocytes and variable amounts of serum. During this late stage only a few of the endothelial leucocytes were observed wandering to or from the lung with a load of pigment or cell debris. These wandering endothelial cells, however, appeared to become loosened from their normal situations and in the vicinity of lymphatic nodes or communicating channels where these cells are prone to localize with their carbon pigment, again assumed their spherical form and took on migratory properties entering into the nearby tissues and scattering themselves in the looser structures. It is an interesting point to note that these pigment carrying cells, ordinarily assuming a latent existence when their cytoplasm has been crowded with foreign particles will assume all the activities of migrating cells when the oedema of the tissues alters the physical properties not conducive to a stationary existence. These cells will then be found to enter the lung alveoli, often appearing as cells which have only recently picked up their carbon load. When, however, the conditions of the experiment, that is, the production of an inflammatory oedema in the lung, are produced in the tissues of an individual with much anthracosis, he will, during the period of his pneumonia and for some time during convalescence, bring up a greater number of these cells in his sputum than are ever obtained during the times when the lung is not involved. We are convinced that inflammatory conditions of the lung tend to reduce the total number of latent pigment bearing cells present in the involved tissues, and in this way somewhat reduce the grade of anthracosis.

A considerable discussion has arisen concerning the proper nomenclature for the pneumonia or pneumonias found in epidemic influenza. From some quarters have come the reports of a true lobar pneumonia, from others a lobular or broncho-pneumonia and others again claim that the reaction is an interstitial pneumonia of varying distribution. It appeared to us that the gross distribution of the lesions is not alone the criterion for a proper appreciation of the inflammatory states which may arise within the lung. I believe it has been amply demonstrated that the pneumonic reactions appearing in different regions of the United States as well as in different countries are not of a constant kind when viewed alone in the light of the gross picture nor are they constant from the standpoint of their bacteriology. We are of the opinion that the earlier phases of the pulmonary reaction are fairly constant in different places and that this constancy is dependent upon the common virus which initiates the respiratory lesion and which then permits a variety of micro-organisms invading as secondary agents. The secondary agents vary with the community and depending upon their nature the character of the reaction differs from that in other places. It has been well demonstrated that in some regions the hemolytic streptococcus is the important organism following the primary injury by the initial virus. In other places the pneumococcus or the staphylococcus or the M. catarrhalis is found to be of primary importance. Up to the present it has not been shown that the influenza bacillus is not the important organism causing the initial reaction and being responsible for the opportunity of secondary invaders leading to such diverse reactions in the lung. In our series we have met with lobar, lobular, interstitial and broncho-pneumonic types. We have not observed a case of the miliary bronchial reaction as described and illustrated by Goodpasture and Burnett and fully investigated by MacCallum. Moreover we have not met with the type of purulent bronchitis as a characteristic lesion preceding pulmonary involvement. The occurrence of pus within the bronchi occurred not early in the pulmonary lesion but later after the bronchi and bronchioles had passed through their stages of acute, serous and hemorrhagic pneumonia and were entering upon their secondary stage with pus production. The pulmonary lesion had long preceded the appearance of pus in the bronchi. We do not hold, however, that such relations between the pulmonary lesion and the purulent bronchitis do not exist for there is evidence that in particular regions this sequence of events was closely observed.

We cannot, however, correlate our findings with the classification of pneumonias as given by MacCallum. His claim for specific types of pneumonia as a sequel to influenza is based upon his statement that “no satisfactory evidence has been brought forward to show that the epidemic influenza is a bacterial infection. It is evidently a general or systematic infection not especially affecting the respiratory tract and analogous in many respects, as Bloomfield has pointed out, to the acute exanthematic diseases.” Thus we are confronted by two schools concerning the nature of influenza. The one claiming that epidemic influenza is essentially a disease of the respiratory system and the other completely denying this.

I am unable to understand the claims which are put forward to substantiate the second view.

The classification of the pneumonias as suggested by MacCallum would be valuable if it could be applied in a practical manner. We find, however, that his description for the pneumococcus-pneumonia hardly coincides with common observations on endemic pneumonia and if the description is to apply only to the pneumonias associated with influenza wherein pneumococcus alone is isolated we find that our own observations do not coincide with this. The picture offered by MacCallum under this heading was reproduced when the bacteriological findings illustrated the presence of organisms other than the pneumococcus or combinations of these. The most characteristic of his description is the one for the streptococcus-pneumonia which when present alone gives quite a unique picture. The picture, however, is to a certain degree modified by the reactions which precede the streptococcus in the lung. Furthermore to offer as a characteristic picture for the influenza infection of the bronchi the presence of a thick yellow pus is hardly complete inasmuch as this exudate appeared only as a stage in the inflammatory process. The intense serous and hemorrhagic response observed early in this type of infection is more unique than the presence of pus which appears somewhat later and which may occur with infections other than the B. influenzÆ. It has long been the hope in pathology to be able to establish by the character of the tissue reaction, the nature of the infecting agent. Up to the present this has been possible only with a very few types of bacteria.

Lung—Stage of Resolution

The removal of the infection and the inflammatory exudate from the lung tissue is accomplished slowly. Clinically the pulmonary process clears up by lysis, and it is quite unusual to have a crisis with the rapid disappearance of the serious manifestations. It is difficult to obtain a clear conception of what takes place in any individual case recovering from an influenza-pneumonia, but if we have an understanding of what may occur in the inflamed lung tissue in any one of the stages or varieties of kind, we may visualize the changing character of the lung condition tending toward the final restoration.

We have previously pointed out that the early stage of influenza-pneumonia is one of congestion, oedema, hemorrhage and more or less leucocytic infiltration, and that this reaction differs materially from that observed in pneumococcus lobar pneumonia. There being no stage of true red hepatization, it has also become apparent that this peculiar primary reaction need not pass into the stage of gray consolidation. Scattered areas in the lung pass from the condition of acute serous and hemorrhagic pneumonia to a type of purulent pneumonia while much of the remaining tissue continues in the state as seen in the early reaction. A certain amount of cellular exudate makes its appearance but not sufficient to lead to a true consolidation. This variety of reaction is present from the fifth day of the pneumonia onwards and may continue with all of its varieties through until the tenth or twelfth day or even longer when recovery from the infection is beginning. Thus the stage of resolution makes its appearance before the inflammatory reaction in the involved lobes has assumed a common character and where we are able to recognize different grades of severity and different stages of inflammation within the same lobe. Resolution taking place in such a lobe has responses occurring in the different parts determined by the nature of the antecedent reaction. We have found that those portions which have not advanced beyond the stage of oedema and hemorrhage may clear up with the disappearance of this early exudate and its infection. In a neighboring portion the purulent inflammation passes through phases differing somewhat from the preceding but also tending toward the restoration of the parenchyma and the disappearance of the inflammation. It would be incorrect to consider the resolution of the early type of inflammatory reaction as an abortive process inasmuch as it is not yet clear whether this serous and hemorrhagic process is not the characteristic inflammation of a peculiar micro-organism or organisms and that when acting alone these bacteria do not in themselves stimulate a further inflammatory response. Hence if it is true that there is a peculiar inflammatory reaction of a non-suppurative and non-fibrinous kind the manner of resolution will differ somewhat from that where these other constituents of the exudate are present. It becomes clear, therefore, that in influenza-pneumonia all of the lung involved in the early peculiar inflammatory reaction need not pass through those stages and reactions as we recognize them in pneumococcus lobar pneumonia.

The resolution taking place in the areas of serous and hemorrhagic pneumonia is accomplished largely by a reabsorption of the fluid, autolytic disintegration of the red blood cells and a certain amount of phagocytosis of red blood cells and their debris. This resolution is quite rapidly accomplished, and the clearing up of such an area may take place in a remarkably short period of time. The leucocytes and endothelial cells which are present with every such reaction become active in phagocytosis of bacteria, and we have repeatedly observed them crowded with small Gram negative bacilli, whose morphology is similar to that of the B. influenzÆ. These areas contain but few bacteria of other kinds. The exudate in the alveolar walls is also simple in character and is readily removed. Slight suffusion of blood, serous fluid, and migrating cells may occupy portions of the alveolar walls during the acute reaction, but these, too, are easily removed and the tissue rapidly resumes its normal character. The vascular and lymphatic congestion again disappear and the tissues which once were soggy return to a normal state without leaving behind evidence of the pulmonary incapacity. The lining epithelium of trachea, bronchi and alveoli is restored by proliferation from the neighboring less injured parts.

If this early stage in influenza-pneumonia is to be compared with the early reactions of endemic pneumonia, it is interesting to note with what ease the resolution may be accomplished in the former, whereas in the latter a further sequence of stages must apparently be passed through before the lung is cleared of its inflammatory products. As we have intimated before, the early exudate in these two types of pneumonia differs very essentially, the one being accompanied by much fibrin and leucocytes which are present only in small quantities in the pulmonary lesion of influenza.

Resolution of the other portions of the involved lobes in influenza is not so easily accomplished. Where a progressive lesion with its development of pus occupying both the air sacs and the tissue of the lung, the outcome of attempts at repair are uncertain. Complete resolution with complete disappearance of the purulent exudate may take place as we see it in many other regions occupied by a similar reaction; and where the purulent response is not accompanied by material damage to the tissue the restoration of the lung is so complete that upon its recovery no evidence is left behind of the former injury, but in as much as the presence of a purulent reaction in the lung is often of more severe grade than this, a certain amount of tissue destruction having been accomplished, the repair does not completely restore the tissue to its former normal state. The purulent lesion, however, is not uncommonly accompanied by minute capillary thromboses, tissue derangement, organic destruction, with even tissue alteration amounting to abscess or gangrene, and it is too much to hope that the lung may be completely restored. Minute abscesses varying from microscopic size to large cavities, several centimeters in diameter, were not unusual in the tissues severely involved in the purulent reaction. Thus in these areas, resolution can be accomplished only by a process of slow organization of the damaged parts with the final production of fibrosis. These fibroses are of variable extent depending upon the initial damage. We have been very much struck with the speed with which this process of organization may take place and the extent of the lung tissue which may become involved in this late lesion. In one of our cases we have evidence of marked fibrosis present on the twenty-third day of his illness. Patches of organization varying from one to four centimeters in diameter occupied the different lobes of the lung. The new fibrous tissue was well developed and the purulent reaction had largely disappeared. The fibrosis obliterated the normal architecture of alveoli and bronchioles, leaving only irregular islands of epithelium which assumed grotesque glandular shapes and looked not unlike a new growth. One of the interesting features of these late fibroses which come to occupy various extents of the lung and bronchial tissues is that the individual after recovering from his acute influenzal lesions again passes, in about his third week, into a stage of dyspnoea with manifestations out of proportion to the physical signs or constitutional derangements which can be determined. The dyspnoea is often the outstanding sign and the patient may die in a state of asphyxia.

We have observed evidence of organization in its earlier reactions taking place in the patches of gray consolidation. This organization of the lung tissue takes place as an interstitial fibrosis and as an alveolar organization. Masses of granulation tissue grow out into and come to occupy the lumen of the air sacs, while in other instances the new growth of tissue takes place mainly in the alveolar walls converting them from thin partitions to thickened and tough structures. In the cases in which a purulent pneumonia was present for some time, and where some of these tended towards repair, this type of restoration with the new development of connective tissue was found. The amount of fibrosis varied very much, and in many instances there was no evidence that obstruction to the bronchioles occurred to a material degree. Hence, although we believe that more or less organization occurs in all of those cases which have passed through a purulent pneumonia, and that a permanent mark is left upon the lung tissue, it is not probable that the amount of involvement and final damage by fibrosis is sufficient to seriously influence the pulmonary respiration. There is, however, a certain percentage of cases in which this organization and fibrosis does involve sufficient of the lung parenchyma and bronchioles to interfere with the pulmonary ventilation.

Where the purulent pneumonia has markedly involved the parenchyma, and particularly where vascular channels both large and small have suffered, some of them by thrombosis, others by a sclerotic thickening, the circulatory disturbance may be sufficiently interfered with to infarct the area. The infarction usually occupies the purulent area itself, and with the complete occlusion of the circulation the resulting necrosis gives rise to an appearance different from that usually seen in pulmonary infarcts. The area may lie in the peripheral portion of the lobe or may occupy deeper parts. The infarct is of a cream-white color, quite homogeneous, and resembles the appearance of a local area of caseous pneumonia. This appearance is brought about through the local purulent consolidation undergoing necrosis. Some of these areas rapidly develop a cavity through liquefaction of the exudate.

The localization of the inflammatory products not only upon the surface of the air sacs but also in the stroma of the alveoli; the interlobular trabeculÆ, and about the vascular channels indicates the intense effect of the virus of this disease. The exudate is largely an indication of the point of action of the irritant upon the tissues, and in influenza with its variety of bacteria in the lung this is not limited to the surface membrane of the air sacs. During this second stage of the reaction the purulent exudate was found occupying all structures of the involved area. Damage upon the component tissues was to be seen in the endothelium of the capillaries, the muscle tissue of the bronchioles and arterioles, the connective tissues and the epithelium. It was seldom that bacteria were demonstrated in the interstitial parts, and it would appear that the damage was the result of their toxins.

Hence, broadly speaking, the end result of the pneumonic process in influenza is far more complex and indefinite than that in lobar pneumonia. Resolution may take place early with the clearing up of the first products of the exudate; or it may be delayed in association with the secondary purulent process which not uncommonly occupies multiple lobes. Where the resolution begins in purulent regions the final outcome is most variable, depending upon the amount of damage which has been imposed upon the lung tissue during the suppurative inflammation, ending either in complete restoration or slight fibrosis of the lung, or passing on to focal scarring of various degrees, sufficient to alter the pulmonary capacity. In other instances the resolution is delayed by the development of abscess, infarct and gangrene. Here the final outcome is determined by the amount of tissue involved in the destructive process, and the persistency with which the infecting micro-organisms attack the local tissues and the constitutional resistance of the individual. Those individuals in whom resolution begins before there is much purulent pneumonia stand the best chance of having the lung return to its normal characteristics.

Pleura

Inflammation of the pleura was a complication which varied in its extent and appearance. It appeared to us that a definite interval lapsed between the development of the lesions in the lung and the appearance of an inflammatory reaction upon the pleural surfaces. Although we have recorded evidence of a pleural reaction in 27 cases, this does not indicate that we have met with that number of pleurisies of clinical severity. In this group we include all gradations of pleural reaction from the merest evidence of irritation and slight dulling of the surface to the cases in which definite and marked inflammatory exudate accumulated within the cavity. In many cases we observed a slight increase in the amount of the fluid present in one or other pleural cavity, while there was little or no macroscopic evidence of a cellular or fibrinous exudate. An examination of the fluid showed the presence of lymphocytes and endothelial cells in small numbers, and sections of the pleural surface at points where a slight dulling of the serous membrane was seen at autopsy showed the presence of a very thin layer of a hyaline fibrin. By taking these reactions as indicative of pleurisy we have recorded 6 cases of acute fibrinous pleurisy, 20 of acute serofibrinous pleurisy, and 1 of acute fibrino-purulent pleurisy.

An increase in the quantity of fluid in the pleural sacs was the most common indication of pleural irritation. The quantity varied from 50 to 500 c.c. of a clear or slightly turbid fluid. Not uncommonly this fluid was blood stained and evidence of superficial extravasation of blood could be recognized directly beneath the pleural membrane. These serous reactions accompanied the early acute stage, while hemorrhage was the accompaniment of the early period of the influenzal pneumonia when similar hemorrhages were found in the lung substance. The pleural reactions were almost entirely confined to the visceral pleura, and only in the very severe responses did we obtain a marked inflammatory reaction with hemorrhage upon the chest wall. Goodpasture and Burnett state that “there is commonly a moderate serous effusion in one or both pleural cavities amounting to 50 or 250 cubic centimeters. The fluid is clear and has the color of blood-stained serum. The pleural surfaces are smooth, shiny and wet, though occasionally a thin, granular fibrinous exudate may be seen by reflected light over limited areas. Often numerous small, red, discrete, or confluent pleural hemorrhages are present over consolidated portions, especially posteriorly on the surface of the lower lobes.” Where organisms other than the influenza bacillus had invaded the pleural sac and had been present for a sufficient time to obtain a reaction, the serous type of exudate observed in the early lesions changed to the turbid type of fluid accompanied by more or less fibrin deposit. There was one case where the intense reaction with fibrin and leucocytes gave rise to a new character to the pleural exudate, a fibrino-purulent pleurisy or empyema.

As we have subsequently learned the pleurisies developing late in the course of the influenza and those which persist after the pulmonary inflammation has passed are prone to be of a purulent kind. There have been a fair number of cases of empyema brought to our attention by the surgical department in the bacteriological laboratory of the hospital, subsequent to the wave of epidemic influenza. If one were to base his finding alone upon observations obtained in the operating room, he would be impressed by the fact that the pleurisy accompanying the epidemic of influenza is of a purulent type. On the other hand, if one were alone to consider the findings at the autopsy table during the five weeks of the epidemic, one would be of the opinion that the pleurisy is of very minor consequence and of a serous type. It is this changing picture which is particularly to be kept in mind. And our experience indicates that during the height of the influenzal lesions of the lung when the pulmonary lesions develop so rapidly that we obtain a pleural reaction closely resembling the inflammatory conditions in the lung and also containing bacteria not unlike the pulmonary flora. Dr. Holman has obtained the influenza bacillus and other varieties from the pleura during these early periods of the pulmonary inflammation. It is more than probable that just as in the infection of the lung tissue where there is a change in the type of the bacteria present, so, too, the flora of the pleura alters in the succeeding stages of the pulmonary reaction. In the late event of empyema we have not observed the influenza bacillus. The majority of the empyemas possess hemolytic streptococci and occasionally pneumococci.

Heart

During the acute epidemic and while the disease was at its height it was remarkable how few cases showed involvement of the heart. It was the common observation that even during intense illness the heart action remained fairly stable and did not indicate an effect by intoxication as might be expected from the severity of the illness. In as much as the majority of deaths occurred within relatively few days of the onset of the severe infection, the type of lesion that would be looked for in the heart would be either bacterial inflammatory products within the pericardium, myocardium or endocardium or toxic lesions of musculature alone.

In our series we have encountered no cases of pericarditis. This lesion in the experience of others has also been unusual, and it would appear that bacterial invasion of this sac is accomplished mainly in the presence of secondary infections localizing in the neighboring pleura. It was not uncommon to find a slight increase in the serous fluid in the sac, but this on no occasion amounted to a hydropericardium. The fluid was always clear and with no evidence of fibrin or cellular exudate. Petechial hemorrhages scattered over the epicardium were noted in seven cases. In the majority of instances these minute hemorrhages were scattered in small numbers over the ventricular walls. In one instance these petechial hemorrhages were also present through the myocardium, suggesting the influence of an intoxication not upon the tissues of the heart as much as upon the finer structures of the vascular channels. This is furthermore borne out in the presence of petechial hemorrhages confined not to one organ, but to various tissues and structures in the body.

More or less cloudy swelling or granular degeneration of the muscle elements of the heart was not uncommon. It was sufficiently pronounced in 12 cases to be readily detected by the naked eye. A lesser amount was also observed in other cases on microscopical examination. In only one instances was the myocardial degeneration of such extent to lead to a definite and recognizable weakening of the musculature. In this instance the autopsy showed a flabby myocardium which was relatively soft and easily broken and in which all the chambers of the heart were decidedly dilated. This was the only case in which we were convinced of a sufficient influence of the toxic effects upon the musculature to permit a stretching of the walls, with failure of function.

In a number of other instances, however, in which there was more or less granular degeneration and cloudy swelling we found that the right ventricle ceased in diastole without, however, the capacity of the chamber being enlarged. We would make this differentiation in speaking of dilatation of the heart. We have met with 11 cases in which the right heart died in diastole, but in which there was no evidence that the right ventricle had been unduly expanded. In four cases there was evidence of an old compensatory hypertrophy of the left ventricle in which the cavity of this chamber was also slightly larger than normal. The lesions in these four cases, however, bore no direct relation to the results from the influenza infection. The appearance of the musculature with moderate grade of cloudy swelling suggested some oedema of the tissues. In the myocardium, oedema is difficult to recognize, and we would not place great stress upon its presence in mild degree.

The microscopic examination of the myocardium showing cloudy swelling gave the usual picture as is seen with a variety of infections. The muscle fibers showed a fine granular deposit in their cytoplasm and the staining quality of the tissue was somewhat altered. The transverse striÆ were less distinct than normal, while not uncommonly the longitudinal fibrils became more evident. Fatty degeneration was not encountered.

In the single case showing a definite and acute dilatation of the ventricles the cause of the myocardial lesion could not be placed at the door of the influenzal infection. This was the case suffering from a secondary streptococcal bacteriÆmia arising in the middle ear. It is more than probable that the streptococcus was the immediate cause of the acute muscle change and weakening. In a number of cases we have studied the tissues of the bundle of His, but we were unable to note any definite change.

It is interesting that the intoxication associated with acute influenza is selective in localizing in certain muscle tissues. We have previously indicated the intensity of muscle degenerations occurring in the abdominal recti. Even in these cases where these striped voluntary muscles were markedly affected the myocardium showed nothing more than a mild or moderate grade of cloudy swelling. We can only account for this in a difference in the constitution of these muscular structures, some being of such composition permitting of the localizing and damage by the unknown intoxicant. It does not appear that the reason for localization in certain tissues is in any way related to the character of the blood supply, nor is it related to the activity of the part.

In three cases we have found an inflammatory lesion of the endocardial tissues. In all of them this consisted of a slight acute verrucose mitral endocarditis. The lesions were very small, consisting only of a fine granular deposit looking like grains of sand localized along the border of the mitral leaflets. In no instance was the leaflet injured or incapacitated. Unfortunately the lesion not being suspected was encountered after the heart had been removed and opened and when it was too late to make bacteriological analyses. This point is greatly to be regretted, in as much as it is of great importance to know whether some distant lesions are induced through the influenza bacillus or its symbiotic flora.

The majority of authors report but little upon the heart lesions in influenza. Many deny that a heart involvement is to be found, a few report an occasional endocarditis. Wallis and Kuskow found more or less myocardial change similar to what is usually described as cloudy swelling. This reaction they point out differs in no way from the degenerations arising from other types of intoxications. Keegan in a series of about 23 autopsies found only a single case with acute dilatation.

Abrahams, Hallows and French had an opportunity of observing over 400 autopsies upon the influenza patients, and they comment upon the infrequency of cardiac dilatation. A slight dilatation of the right ventricle was seen in a few cases, and in no instance did they find pericarditis or endocarditis. They comment upon the heart condition as follows: “The most remarkable feature about the heart is the general absence of dilatation. In quite a large proportion of cases there has been no trace of dilatation; in a fair number of others there has been some dilatation of the right side, but this has seldom been extreme, perhaps enough to cause the apex of the heart to be formed about equally by right and left ventricles. Most often the heart has appeared of normal dimensions and the apex has been formed entirely by the left ventricle. This absence of dilatation accounts for the clinical absence of orthopnoea.” In direct contradiction to the above findings, the Advisory Board to the D. G. M. S., France, report the findings in 30 autopsies of clinical influenza. Twenty-nine of these 30 cases showed dilatation of the heart, chiefly of the right side, but very commonly of the left side as well. Twenty-one showed myocarditis and two endocarditis. In this report it is stated that these patients showed evidence of obsolete tuberculosis. It is possible that the condition of the patients and the presence of an unusual complicating infection led to the high incidence of cardiac involvement. The figures in this last series are much too high when compared with the frequency of heart involvement as found by the majority of other investigators.

A number of heart lesions not resulting from influenza were observed. For none of them was there an antecedent history, but in some cases the condition may have had an influence in causing accessory cardiac embarrassment. One case had a chronic interstitial myocarditis of the rheumatic type, three had mild grades of chronic sclerotic mitral endocarditis, one a bicuspid pulmonary valve and three showed old pericardial adhesions, one of them having a complete obliteration of the sac. The foramen ovale was patent in six of the hearts.

Arteries

The arteries in these young adults were remarkably healthy, and in none of them did we observe the characters of arteriosclerosis or leutic lesions. On the other hand, evidence of superficial fatty streaks lying in the intima of the aorta and some of its large branches were not uncommon and are believed to have had a relation to the acute infection of which they died. In only four cases in the series of 32 autopsies was evidence of these fatty streaks wanting. In about one-half of the remaining number these fatty streaks were only slight or moderate in extent, while in the rest of them these lesions were particularly prominent and striking. They formed linear markings on the posterior wall of the aorta, aggregating with particular prominence about the intercostal arteries. The anterior wall was quite free from them. The greater extent of these lesions lay in the descending thoracic and was less marked in the arch and the abdominal aorta. At times these fatty streaks were found to extend into the large vessels of the neck and into the intercostal arteries, and they were also found in the coronaries of the heart. It was uncommon to observe their presence in the arteries of the abdominal viscera.

This type of lesion has been discussed from the standpoint of its etiology and its possible bearing upon true arteriosclerosis. Some believe that the frequency of its finding in autopsy material suggests the non-importance of its presence. This we can hardly agree with. It is true that the presence of these lesions does not materially incapacitate the aorta in acting as the main channel for the distribution of blood. The lesions are quite superficial in the intima and cause but little elevation on the surface. The amount of roughening which the intima presents to the blood is not great. Nevertheless, the presence of these fatty streaks is an index of the disturbed metabolism of the cholesterin products of the body. Under certain conditions they make their appearance when there is a true hypercholesterinemia such as is readily produced in the animal experiments by feeding cholesterin. Under these circumstances the various tissues of the body, including the adrenal, the corpus luteum, the spleen, liver and arteries, all participate in localizing cholesterin in the form of cholesterin-ester in peculiar cells which have been termed cholesterin-ester phagocytes. It has been shown that cholesterin metabolism is quite readily altered in the human and that the blood content will vary from the normal. In chronic kidney disease, pregnancy, diabetes, chronic heart disease and arteriosclerosis the blood cholesterin rises, while in many of the acute infectious diseases the cholesterin in the blood is materially diminished. It is particularly in these latter cases where fatty streaks of the intima are prone to occur. Hence in human pathology we more often meet with the development of fatty streaks of the intima associated with a hypocholesterinemia than with a hypercholesterinemia.

The fatty streaks of the intima of the aorta to which we are referring are lesions quite aside from true endarteritis as well as atheroma. In naked eye appearance the lesion is of a fatty nature and suggests atheroma, but it differs from this well-known lesion in the fact that the fatty materials, cholesterin-esters, are contained within cells which are of uniform type and have no reaction in their immediate vicinity. True atheroma may occur in definite levels of the intima, most commonly in the deepest portion, and is characterized by the fact that we are dealing with a variety of fatty materials, neutral fat, fatty acids, soap, cholesterin-ester and free cholesterin which lie between the tissue cells forming a detritus following a process of true degeneration. It is possible that some of the superficial fatty streaks do give rise to a small atheromatous area by death of the cells which primarily contain the fatty substances. Most commonly, however, the fatty streaks do not progress directly to atheroma but may entirely disappear, as we have seen it occur in our experimental animals. At other times these fatty streaks are followed by a slight thickening of the surface of the intima so that the resemblance to early endarteritis is obtained. We do not believe that these fatty streaks in themselves lead to the chronic nodular thickening of the aorta, but that other factors giving rise to a low grade inflammatory reaction must be present.

There appears to be a relation between the development of these fatty streaks and the altered cholesterin metabolism, brought about by pathological change in the blood, adrenal cortex and it may be in the liver. It is under these conditions where these tissues are altered particularly by bacterial toxins in a process of marked cloudy swelling that these intimal fatty streaks arise. Analyses in other diseases have shown that such organic changes lead to a diminution in the cholesterin content of the blood, while at the same time there is neither an increased intake nor an excessive output. It would appear that certain types of tissues and cells are stimulated into activity to become depots for the cholesterin which is not being properly handled by the adrenal and other organs. These cells in the intima which become active in taking up cholesterin-esters are types of endothelial cells whose origin is not entirely clear. In these lesions it is observed that the most superficial cells of the intima do not show an overloading with the fatty compound, but that the cells active in absorption lie at a level slightly beneath the endothelial lining and form colonies as if arising through active division of cells which are present in these parts. Active migration on the part of these cells is not to be observed. They do not appear to wander far from the location where they are found during the acute process. The plaque may enlarge by proliferation and thus enlarge the extent of the involved area. We have failed to find, however, that these cells migrate into the lowermost portion of the intima or into the media. The possibility that these cells do arise from the endothelium lining the blood vessels has, up to the present, not been excluded. If such is the case, the cells appear to adopt a function which is not commonly observed in normal arteries nor present in the endothelial cells lying immediately above the fatty plaque.

We have searched various arterial systems in the cases of acute epidemic influenza for inflammatory lesions lying in the adventitia and media. These, up to the present, we have not discovered. Some years ago a number of French authors reported the development of acute non-suppurative influenza lesions in the outer coats of arteries which at times had aneurysm as the outcome. These cases, however, occurred during non-epidemic periods, when the type of influenza of which the patient suffered was quite different from that seen in pandemics. As far as we know none of the reported cases of arteritis and aneurysm occurring under these conditions has shown the presence of the influenza bacilli in the arterial lesion. It is possible that sporadic influenza has complicating secondary infections which are of importance in localizing in the arterial wall.

Occasional reports have been made upon the occurrence of thrombosis immediately following an attack of influenza. These thromboses have occurred in diverse regions, the brachial, femoral, the mesenteric, and other arteries. It is possible that the development of the deep hemorrhagic lesions of muscles in the extremities are associated with thrombosis. It is impossible, however, to demonstrate within such blood masses the presence of thrombosed vessels which had preceded the hemorrhagic state. It was, however, possible to demonstrate capillary thromboses through the lung and in the submucosa of bronchi and trachea. In these instances the damage to the vascular walls was brought about by the action of the infection immediately surrounding them, and was not associated with a process beginning within the lumen of the channel. The type of thrombosis within the lung to which we have referred in a previous discussion is interesting in that it does not show the usual type of fibrin clotting, but in place of fibrin threads a gummy homogeneous material is deposited upon the vessel walls within which the red blood cells soon undergo dissolution. It would appear that these thromboses within the lung are dependent upon a toxic action on the vessel wall and its plasma content.

Thromboses within venous channels are met with more often than in arteries. The veins of the lower extremities are most frequently affected, and yet amidst the many cases of influenza it is an unusual occurrence. The various thromboses of larger vessels usually occur as post-influenzal complications rather than as accompaniments of the acute disease. It is possible that factors other than those present during the acute stage play an important part, and that the virus of influenza is not directly the cause of the thrombosis.

Lymphatics of Lung and Mediastinum

One of the prominent reactions which was almost constantly present as the inflammatory reaction involving the lymphatic system of the chest. The lymph glands within the chest responded to a marked degree in hyperplasia and commonly showed enlargement quite out of proportion to what is usually observed in lobar pneumonia. These reactions were in direct relation to the inflammatory processes of the lung and appeared to be involved in proportion to the inflammation occupying the tissues drained by them. Elsewhere in the body the lymph glands responded but slightly, and often no change was observed in the lymphatics of the abdomen, axilla and lower extremities. The systemic intoxication thus had no effect upon distant lymph glands, and even the presence of micro-organisms in the circulation did not appear to cause responses in these tissues other than in the neighborhood of the chest. Within the chest the lymphatic system became involved through the presence of the various bacteria migrating along the lymphatic channels as well as through its activity in removing products of inflammation.

The response of the thoracic lymphatics, including those within the lung and mediastinum, is observed in all stages of pneumonia. But in epidemic influenza the reaction was much more prompt, appearing in the early stages and rapidly developing tissue changes along the channels and in the lymph nodes. The lymph channels during the period of the early serous pneumonia became dilated and filled with fluid with relatively few cells. The stroma immediately surrounding became oedematous, so that in the gross specimen the connective tissue between the lobules of lung were sometimes easily seen as gray strands. At this time this tissue was not increased in quantity and did not project above the level of the cut lung. The fibrous tissue remained soft and pliable, but formed quite wide strands. When the pulmonary reaction became hemorrhagic, red blood cells, leucocytes and large mononuclears were found mixed with the fluid in the lymphatics. We had no way of determining the direction of the lymph flow from the pulmonary tissues, but it was assumed that as there was no excessive loss of serous fluid from the lung and the lymphatics beneath pleura into the chest cavities that the fluid was draining through the channels lying about the bronchi and vessels. The further evidence of the direction of flow was seen in the rapid and comparable responses which occurred in the lymph glands along these routes. The glands about the bronchi and at the hilus became enlarged, red and succulent. The glands were often two and one-half centimeters in diameter. Their capsule was thin and stretched and the gland was quite soft. Many of them when cut open were almost diffluent.

This acute lymph hyperplasia occurred in 30 of our cases. It is impossible to indicate any particular type of infection as being responsible for these lymphatic lesions. The nature of the bacteria present in these 30 cases differed quite considerably: 25 showed influenza bacilli, 15 pneumococci, 18 streptococci, 8 M. catarrhalis and 17 staphylococci. In as much as the pulmonary reaction was fairly constant in certain characteristics in all of our cases, and as we believe that the influenza bacilli were the very important factor in these reactions, it would appear that the lymphatic responses are only a part of the general inflammation of the respiratory organs. Comparison can also be made of the character of the lymphatic changes with that occurring within the pulmonary tissues. The lymphatics were filled with fluid which dilated all the available sinuses; the lymph nodes were oedematous and within them the reaction often had numerous small hemorrhages.

The lesion within the lymph nodes following the early serous inflammation was of a non-suppurative kind. The lymph follicles lost their outline, and the lymphocytes were diffused through the stroma so that no recognition of the germinal centers could be found. The dilated sinuses within the lymph nodes were filled with large mononuclear cells, of the type of endothelial cells, along with some lymphocytes and leucocytes. Subsequently the leucocytes increased very materially so that the lymphatic fluid became purulent. Smears obtained from larger lymphatics showed leucocytes and varieties of bacteria. This was particularly true in those cases where the pulmonary lesion had itself become purulent either localized in a patchy pneumonia or with lobar involvement. Under these circumstances focal areas of purulent infiltration were found within the tissues of the gland occupying the regions of the former follicles and leading to necrosis or abscess. Where such purulent reaction and abscess formation were found within the lymph nodes there was remarkably little reaction in the tissues of the immediate vicinity. No attempt at the development of a pyogenic membrane or granulation tissue was observed, though this probably does take place in the cases recovering.

In only one instance did we observe the development of the peculiar fibrosis along the lymphatic channels where the freshly cut section of lung reveals prominent and raised demarcation between the lobules. This response has been described by MacCallum as unique for the streptococcus inflammation of the lung. The character of the exudate within the lymphatics with many mononuclear cells and blood is not to be considered singular for the influenza pneumonia. It has been found that in ordinary lobar pneumonia, as well as in the pneumonia following measles, the early pulmonary reaction is accompanied by the dilatation of the lymphatic channels along the bronchi, containing serous fluid, mononuclear cells, blood and leucocytes, while occasionally thrombosis entangling bacteria is also encountered. It would seem, however, that the lymphatics in epidemic influenza can more readily recover their normal character when a streptococcus infection is wanting.

In the late purulent lesions of the lung we have encountered dilated lymphatic channels whose yellow contents could be recognized by the naked eye. At times this could be followed for short distances along the bronchi as narrow yellow cords, or when cut transversely appeared as small dots close to the bronchi or vessels. On pressure small droplets of pus may be evacuated, or again where fibrin has led to a coagulation of the exudate a yellow plug can be withdrawn from the channel. These small plugs resembled the thick exudate seen within the bronchi and often were misleading when first viewed. The distribution of the purulent lymphatic masses was most irregular occupying only local or patchy fields in the lung, particularly associated with the purulent confluent pneumonia. In one instance such a lymphatic appeared to be associated with the development of a small abscess lying close to the bronchus.

Too much stress cannot be placed upon the importance of the lymphatics in all forms of pneumonia. They play an important role in the drainage of the lung during inflammation. In the normal lung we hardly appreciate the lymphatic distribution except in our observations upon anthracosis. But even under these conditions when much carbon is deposited in conjunction with the lymphatic system we do not gain a true appreciation of the activity of the lymph channels and nodes during an acute process. Bacteria may be demonstrated in acute infections of the lung within the fluid and cells of the lymph channels. Less easily may we demonstrate bacteria in the lymph nodes under similar conditions, although when abscess has occurred their presence is readily recognized. The transport of bacteria is accomplished not only by a passive migration of micro-organisms in the fluid as it drains from the lung, but organisms are also found within the leucocytes as they travel with the current. Only occasionally have we demonstrated bacteria within the wandering large mononuclear cells, although we have observed them in a few instances within the cells lining the sinuses of the nodes.

Whether the inflammation of the pleura is directly related to the involvement of the pleural lymphatics we have not been able to determine. In our series of cases pleurisy has not been a prominent feature of the disease, and in many instances the grade of involvement was so slight that it was not easily recognized by the naked eye and showed only a slight reaction microscopically. That the presence of bacteria within the intricate plexus of lymphatics beneath the pleura may be responsible for the development of an inflammation of this membrane may well be the case, and in this way simulate the mode of transmission of the infection as seen in lobar pneumococcus pneumonia and in the streptococcus type of infection.

Abdominal Viscera

The lesions occurring in the abdominal viscera were of less importance than those within the thorax. In none of the cases of the epidemic was the intestinal type of the disease, described in previous years, encountered. The changes found in the various viscera were concomitant with evidences of intoxication as observed clinically or at autopsy in other regions of the body. We found no evidence that the bacteria of the disease localized in the tissues of the abdominal viscera, and we were led to believe that the alterations in morphology and function were the result of diffusible toxins. The action of these toxins was either upon the parenchymatous cells of the organs, as in the liver and kidney, resulting in granular degeneration, or upon the capillaries with the development of petechial or diffuse hemorrhage as was encountered in the stomach, intestines and bladder. The absence of definite localized inflammatory processes in these distant tissues, including the abdominal lymphatics, speaks against the probability of a bacteriÆmia playing an important role in the disease. That transient bacteriÆmias by the influenza bacillus do occur has been repeatedly demonstrated, and that the organisms associated with this bacillus may also enter the blood stream has likewise been found. But these states are accessory to the disease, and must be viewed as complications rather than the rule. Hence the occasional observations by some, of bacterial inflammatory reactions in liver and kidney must not be considered a part of epidemic influenza, for in many cases it is wanting. The majority of lesions of the abdominal viscera probably arise through the action of the unknown toxin in the blood.

In the stomach and intestines the lesions were of two kinds, (1) hemorrhage and (2) erosions. Petechial hemorrhages were present in the stomach 15 times, in the intestines 4 times. These small dots of blood extravasation, lying in the mucosa and submucosa, differ in no way from those observed in other acute infections and intoxications, save that the tendency for the leakage of blood into the lumen of the viscera was more pronounced. Often we could observe the presence of free and more or less altered blood in the stomach and intestines, and in 12 cases the amount was considerable, sufficient to be spoken of as melena. It is probable that the oozing of blood takes place not only from the areas visible to the eye as petechial hemorrhages, but also from the more normal-looking mucosa of stomach and bowel. The tendency to hemorrhage was not necessarily accompanied by visible alterations in the epithelial layer of the mucosa, though at times erosions were found. When hemorrhage could be observed, the extravasation of blood occupied the superficial layers of stroma, causing a separation of the tissues beneath the epithelial layer. At times the submucosa was also infiltrated, and in one instance the musculature. The lesions were isolated and sporadic, but always about small capillary loops. It appeared to us that the damage was primarily upon the vascular tissues and particularly upon the endothelial walls of the fine channels. Inflammation was not present, and the hemorrhage was more or less passive—that is, a slow oozing rather than acute hemorrhage by rhexis.

The second type of lesion of the gastro-intestinal canal was erosion. This was of the nature of a defect in the mucosa, usually multiple, small and well circumscribed. The tissue loss was superficial. In their appearance these lesions were similar to those encountered in these parts in other infections, and also as described by McMeans in experimental infections of animals. The erosions appear to arise in a process of bland necrosis, limited in the periphery by healthy tissue and not tending to enlarge. It is probable that these erosions are associated in their development with the petechial hemorrhages, being a sequel to the vascular disturbance of the mucosa and subsequent digestion of the injured tissue. Multiple lesions of the stomach were found 10 times and twice in the intestine. The largest was 1.25 cm. in diameter. They are more common on the posterior than anterior wall, and usually toward the lesser curvature. It is probable that these defects are limited in their progress and heal readily.

The changes occurring in the liver were not of striking account. Cloudy swelling was observed 13 times, usually of moderate grade. The usual appearances with enlargement of the organ, bulging of the parenchyma on section and a dull gray cut surface were all that could be found. The one case with icterus was the only one in which the natural discharge of bile from the liver was interfered with through the swelling. Even in this case the obstruction to the outflow of bile in the small channels was not demonstrable in the microscopic sections, nor was there evidence of unusual bile staining of the liver-points suggesting the possible origin of the icterus in an unusual hemolysis. On no occasion did we meet with recent inflammatory reactions in the gall bladder or bile ducts, and we have no evidence that the organisms of the infection are discharged from the body by these routes. The cloudy swelling of the liver was accompanied by slight oedema of these tissues in seven cases; and in six instances focal necroses were observed. These focal necroses were similar in appearance to those seen in typhoid fever, but were much less frequent in the tissue. Only careful search revealed isolated pinhead gray dots with depressed centers. They were most commonly in the mid-zone of the lobule, and in the early stage were without inflammatory reaction. Subsequently, leucocytes infiltrated the area, but not in an amount to form pus. Bacteria were never demonstrated in the areas of focal necrosis. Four cases showed old adhesions about the gall bladder and in one a gall stone was present.

Lesions of the pancreas were not encountered. In a few cases the lymph glands about the head of the pancreas were slightly enlarged.

The spleen showed relatively little reaction and in only two cases was it enlarged. Fourteen times a diagnosis of acute splenitis was made on examination of the gross specimen. This diagnosis rested upon the finding of a swollen spleen with tense capsule and with a dark bulging pulp. The Malpighian bodies were usually in part or completely obliterated, though in a few instances these grayish nodules seemed even larger than normal. These spleens contained an excess of blood within the pulp. In one case several isolated areas appeared hemorrhagic as if a local rupture of the tissues had occurred. The microscopic examination of these specimens showed mainly a marked congestion of the sinusoids, a diminution in the size of the lymphoid corpuscles and some increase in the number of leucocytes within the blood spaces and reticulum. Only occasionally did we observe a proliferative reaction of the large mononuclear cells lying in the reticulum. This proliferation was not sufficiently marked nor uniformly present to be considered as characteristic. We did not find abnormal deposition of blood pigment indicating an unusual destruction of red blood cells within the spleen. It is interesting to note that 5 of the 32 cases showed obsolete miliary tubercles in the spleen.

Our analysis of the changes occurring in the kidney bore out the clinical findings observed in the wards. Like in so many acute infectious diseases urinary changes were commonly present. These are in part dependent upon systemic changes in the metabolism of tissues and not entirely the result of renal lesions. In acute epidemic influenza there was no common characteristic in the urinary output. The amount excreted in 24 hours was usually diminished to a small extent, the color was darker, the specific gravity slightly increased, as well as the total solids. There was no marked change in the total quantity of output of any one of the constituents as far as they were analyzed by us. Albumin was present in the urine in variable amounts and in the more severe cases casts were also present. There was only one case in which the quantitative output was much diminished and where some fear was entertained of development of acute uremic manifestations. This individual, however, died before these made their appearance and before there was any evidence that the retention of waste products was causing definite clinical symptoms.

In 30 cases coming to autopsy more or less cloudy swelling was to be observed in the kidney. This reaction varied from a very mild swelling and granular degeneration of the tubules of the cortex to a decided parenchymatous degeneration with loss of nuclear structure and erosion of some of the cells lining the tubules. The convoluted tubules were always most markedly involved. Occasionally this tubular degeneration was accompanied by a desquamation of the lining cells of the glomerular capsules. We were, however, unable to recognize an acute inflammatory reaction in the interstitial tissue or in the glomeruli in any of the cases, except the one which had developed a streptococcus bacteriÆmia as a sequel to an otitis media. The kidney lesion reminded one very much of the toxic lesion which is observed in the kidney in typhoid fever. Differing, however, from the latter there was a variable congestion of the fine vessels associated with the cyanosis which was present in a certain percentage of these cases. At times the kidneys were quite wet with blood from the venous engorgement.

The lesions in the kidney were of a toxic type and did not resemble reactions following the presence of the bacteria in the stroma of the organ. In the majority of instances in other diseases where bacteria themselves locate in tissues we are able to recognize focal lesions of acute necrosis or inflammation. In epidemic influenza where a variety of micro-organisms within the lung are able to reach distant structures in a bacteriÆmia, we would, because of their type, expect to find inflammatory reactions of a definite kind. The absence of such reactions is very suggestive that the bacteria do not commonly localize in the kidney, but that their toxins alone affect it during its elimination. We have also entirely missed the finding of any vascular lesions in the renal system. Neither degeneration nor inflammatory reactions of any of the coats of the blood vessels could be distinguished.

The partial incapacity on the part of the kidneys must, therefore, be viewed as a complication resulting from the effect of a diffusible toxin reaching them by the blood stream. The damage performed in this manner may be quite extensive upon the secreting tissues of the tubules leading to an increased or decreased output of the urinary constituents. Because of the nature of the lesion, it is probable that the kidney damage incurred during the acute epidemic influenza is only temporary and not permanent. Tubular degeneration is readily repaired, and in the absence of an inflammatory reaction in the interstitial tissue or the glomeruli avoids the development of a permanent mark or derangement in the system. This is as we find it in typhoid fever.

In two cases we observed very interesting lesions in the bladder. These two individuals during life had been excreting markedly blood-stained urine for some days preceding death. In the one case the hemorrhage was so marked that on standing, about one-tenth of the urine was composed of sedimented red blood cells. It was assumed that the hemorrhage was of kidney origin until the autopsy revealed a simple cloudy swelling of the kidney associated with a hemorrhagic state of the submucosa of the bladder. In both cases the posterior wall of the bladder was heavily infiltrated with blood so that the mucosa was raised from the surface and the prominent folds showed a superficial erosion with small points of greenish necrosis. This bladder hemorrhage was concomitant with hemorrhagic foci elsewhere in the body, pericardium, pleura, stomach and intestine. Alone in the bladder however, the hemorrhage formed a distinct mass and allowed a considerable escape from the lesions on the surface. These areas of hemorrhage were not infected and showed no local inflammatory reaction. They also appeared to be toxic in origin and resembled the hemorrhages occurring in the muscles of the abdomen.

Changes in the adrenal gland were noted in 14 instances. In all of these there was the picture of what is commonly known as cloudy swelling of the cortex and, in addition to this, in three cases small petechial hemorrhages were observed. The so-called cloudy swelling of the adrenal consists largely in a loss of the bright golden appearance of the cortical tissues accompanied by soft oedematous swelling. The tissues change color to a brown or clay color, and it is not uncommon to observe that the inner zone of pigmentation is more diffuse. There is no sharp demarcation between the layers of the cortex. With this alteration in the outer structure of the adrenal, the medulla not uncommonly appears smaller. This change is more apparent than real, and we have not been able to observe any definite lesion in the nervous portion. At times we believed that the inner tissue appeared more cellular, but it was not possible to determine any specific alteration in the cells.

The changes in the adrenal cortex are comparable to those observed in typhoid fever. The analyses of these tissues showed that the cells were almost devoid of cholesterin bodies and few doubly refractile globules could be demonstrated. This change in the adrenal is by no means specific for any acute disease, it being found in many of the severe infections. We regret that systematic analysis of the blood serum in these cases was not made to determine the cholesterin content. If the comparison bears out with typhoid fever, we would expect to find that the quantitative cholesterin of the blood is diminished. Some importance attaches itself to the study of the cholesterin metabolism, particularly in regard to the development of the peculiar fatty streaks which develop in the aorta and other arteries during these acute infections. It has been claimed that in the human these streaks bear an analogy to those produced in the experimental animals and that the arterial lesions are associated with an altered activity on the part of the adrenal cortex in handling the cholesterin compounds. In influenza there is evidence that the adrenal does not function in a normal fashion and that the storage of cholesterin-esters does not take place. From this, however, we cannot conclude that the blood content is increased, and, in fact, it is more than probable in comparing the other reactions of the disease that it follows the changes as seen in typhoid fever where the blood content of cholesterin is lowered. In this way comparison with the experimentally produced arterial lesions in animals is not clear, in as much as in the experimental work a true hypercholesterinemia was induced. Nevertheless it is possible that with the abnormal function on the part of the adrenal the cholesterin materials are made more available for absorption by other tissues and that a true hypercholesterinemia is not necessarily a constant factor, even with the abnormal accumulation of these substances in the intima. It may well be that the normal activity of the adrenal is related to the presence of toxins in the circulation and an attempt by mobilizing cholesterin to diminish the activity of these harmful substances.

The recent epidemic of influenza has afforded a series of interesting autopsies in view of the very extensive and peculiar involvement that occurred in the lungs of the cases examined. Ordinary lobar pneumonia, as we know it, was not observed, although it must be said that the lungs many times exhibited a consolidation of a lobar distribution. The usual dry granular lung of the more common pneumonia was absent, and in its stead a most unusual series of pictures was observed in the several cases. A common feature of all cases was the oedema of the lung tissue, which in the majority of instances contained such an amount of fluid that it ran freely from the cut surface in almost unlimited quantity. This fluid varied in its color and consistence depending upon the age of the process. In the very early cases the lungs were boggy, very congested, and a thin serosanguinous fluid poured forth from the cut surface. It actually appeared as though the fluid within the tissue was under considerable pressure. At times blotchy deep red hemorrhages occurred in the lung substance, and hemorrhages of a bright red color were not infrequent in the pleura. That the circulation of the lungs was much embarrassed was often prettily demonstrated by the dilatation of the fine capillaries and lymphatics beneath the pleura. These small vessels stood out prominently as a meshwork more or less outlining the areas supplied by them. Not only was the peculiar consolidation in lobar arrangement, but also in many cases was there evidence of a lobular distribution. Even in some cases where the entire lobe was consolidated the cut surface presented a peculiar lobulation with patches of lung tissue projecting above the general surface. The wet trabeculated structure of the lung in this stage did not give the impression of true red hepatization, but rather a structure resembling spleen and at times a meaty, compact, glassy picture not unlike thyroid.

As the process advanced the appearance of the lung changed from deep red to yellowish red and finally to a quite yellowish gray color, still retaining, however, the very moist characters. The fluid found in the lung changed its consistency from the thin red type to a sticky, glairy variety which could be pulled out in long strings. It was noted that the change in the character of the fluid was accompanied by similar changes in the lung structure, advancing in two cases to abscess formation of a grape-bunch type. Here there was a rather extensive necrosis and cavitation of lung substance in communication with the bronchioles. However, there was also marked softening and necrosis of lung in a number of cases where abscesses did not develop, but the lesion was so advanced that the lung substance was almost diffluent. An accompaniment of these advanced cases were irregular yellow islands which appeared beneath the pleura. At times they reached the size of a circle 2 cm. in diameter and were slightly raised above the surrounding pleural surface. When these were opened they were found to be areas of softened lung substance. This reaction was so extensive in some lungs that it resembled to a degree the appearance of a caseous pneumonia. However, the former process appeared to be brought about by the interference with the lymphatic drainage, as it was not uncommon to see engorged yellow channels beneath the pleura as well as enlarged lymph nodes at some distance from the hilus. Another feature of the advanced cases were the plugs of ropy yellow material which were contained within the bronchioles, while in the early cases the bronchi and bronchioles showed intense congestion of the mucosa with blood-stained fluid in their lumina.

Of the more unusual reactions observed in the lungs an infarct was found occupying a considerable part of the lower left lobe in one case. There was a marked softening of the lung tissue with reddish, mucky-looking lung substance arranged about small irregular cavities. This reaction extended into the lung for a distance of 4.5 cm. Bordering close on these softened areas there was a dry mottled yellowish gray and deep red lung tissue. Surrounding this area again were noted a number of small blood vessels in which there were found yellowish granular plugs. One plug in a vessel was found at a distance of 3 cm. from the base of the lobe, and another was found at a distance of 8 cm. from the apex of the lobe. On further examination it was observed that the base of this softened area was situated on the pleural surface and that the apex was directed inward about a distance of 6 cm. from the pleura. Bathing the cut surface there was a glairy and very sticky material of a reddish yellow color. Near the apex of this softened area in the lung there was found a vessel about the size of a goose-quill in which there was a grayish yellow granular plug. This plug was adherent to the vessel. Within the small bronchioles there were plugs of a soft yellowish brown material. The striking feature in addition to the softening of the lung in a number of places was the glairy material of a sticky nature which bathed the cut surface. A white infarct was present in the spleen. The lung described above as well as another showed gangrenous change. In the second of these two abscesses had formed, and there was a communication between the lung and pleural cavity in which there was a large amount of sanguino-purulent fluid and a pyopneumothorax.

In a description of these reactions it must be added that the early and late changes were not always observed independently, but in most cases occurred together, giving the lung a peculiar mottled red and yellow glassy appearance. More frequently the congested oedematous reaction was observed singly, while the purulent alteration usually was in combination with the former type. The acute serous pneumonia was noted 13 times, 6 times in combination with the purulent reaction and 7 times alone, while the acute purulent pneumonia was found in 9 cases, 3 times alone and 6 times with an acute serous process. In all but 3 of 18 cases there was evidence of a bronchial distribution. Two of these three cases showed a massive oedematous lung with in one case an extensive hemorrhage, while the third presented an advanced purulent reaction with marked necrosis and softening. An acute bronchitis which varied in character from a hemorrhagic to a purulent one was present in all the cases. The reaction observed within the bronchi in the individual cases corresponded closely to the picture found in the lungs.

In all cases except one there was an exudate in one or both pleural cavities. A serofibrinous pleurisy was noted in 11 cases with, in 2 of this number, a fibrino-purulent reaction present in the opposite pleural cavity, while fibrino-purulent pleurisy occurred alone in 6. In 6 cases pleurisy occurred on one side only with the incidence equally divided in each cavity. Both pleurae were involved in 9 cases. Seventeen of the 18 cases showed both lungs involved. One case was an individual who had had clinical influenza and during convalescence developed gangrenous colitis and acute ascending myelitis which terminated fatally. B. influenzÆ was isolated from the bronchioles in the lung of this individual.

The reaction of the body generally was evidenced by a widespread distribution of petechial hemorrhages over serosal and mucosal surfaces. However, certain other important lesions were noted such as one acute vegetative mitral endocarditis, two acute serofibrinous pericarditis, three cases in which focal necroses were prominent in the liver and two examples of infarct of spleen. Further, there were four cases of slight dilatation of the right heart. The liver was usually swollen and oedematous and the spleen presented evidence of an acute reaction, softening and reddening of its pulp with at times slight enlargement.

As evidence of the virulent character of the infection from which these patients suffered, there was not only present in the lung a peculiar hemorrhage and purulent process, but also a more or less widespread distribution of hemorrhages in other parts of the body. The gastro-intestinal tract was most affected with the stomach showing petechial hemorrhages in 17 of 18 cases and the small intestine in 15 of the same number. In the gastric mucosa of three cases there were definite erosions, while in two instances the duodenum presented an intense oedematous and hemorrhagic appearance of its mucosa. Further hemorrhages were observed on one occasion each in the mesentery and in the mesenteric and retroperitoneal lymph nodes. In the latter the mesenteric glands were so distended with hemorrhages that a soft pulp spurted out when the glands were sectioned. Next in order of frequency, hemorrhages were noted 9 times in the pleura, 8 in the pelvis of the kidney, 6 in heart muscle and 3 each in pericardium and bladder. In one case of widespread distribution of petechial hemorrhages there was a massive loose hemorrhage into the lower recti abdominis. Further another case showed a large amount of a blood-stained fluid in the peritoneal cavity.

Summary

In the analysis of the cases of acute epidemic influenza two important features of the disease present themselves, (1) a marked systemic intoxication with localized manifestations in certain organs, and (2) inflammatory lesions of the respiratory tract. These manifestations present themselves both to the clinician and to the pathologist, and to each they have demonstrated their importance in the disease. The pathologist not in touch with the clinical manifestations of the toxÆmia has more closely linked the occurrence of these two factors with the actual findings in the cadaver. But there are those who look upon these factors as separate and distinct, viewing the toxÆmia as an individual process and as illustrating the uninvolved influenza, while the inflammatory reaction of the respiratory tract is taken to be a complication arising through the activity of secondary invading organisms. This is the view held by MacCallum, who compares influenza with the acute exanthemata wherein the respiratory lesions are but secondary to the production of a lowered resistance and an invasion by a variety of bacteria. Such confusion presupposes an undetermined virus for influenza. In confirmation to such views we have the reports upon a filterable virus. Up to the present, however, the latter has been on insecure grounds.

It would appear to us that, as has been discussed by Dr. Holman, the case against the B. influenzÆ not being the important causative agent has not been proved. The demonstration by others of a potent toxin from the B. influenzÆ cannot be overlooked, and although the actual disease has not been reproduced in animals, there is evidence that this toxin will induce acute degenerations in various tissues. Furthermore, the in vitro symbiotic relation demonstrated for the B. influenzÆ with other organisms, as the pneumococcus, streptococcus, staphylococcus pyogenes aureus and M. catarrhalis, gives ample support to the claim for a similar symbiosis in the human tissues. The evidence for the important primary relation of the B. influenzÆ to epidemic influenza is such that we cannot disregard it—at least, not before we can produce some definite positive evidence that another demonstrable virus precedes it and produces those constitutional effects which initiate the remaining sequelÆ.

We must agree with Christian in the statement that all cases dying during the acute stage of epidemic influenza have inflammatory lesions in the respiratory tract and largely in the lung (pneumonia). It is difficult to conceive of a disease comparable to the acute exanthemata, which beginning as a separate and distinct process ends fatally within 48 hours with a pneumonia which is claimed to be secondary.

Epidemic influenza is an acute infectious process of the respiratory tract, usually localizing in the upper respiratory system, but often and in a fairly constant percentage of cases extending into the lower portion of the same system and causing a type of broncho-pneumonia. Accompanying the initial invasion there is a marked systemic intoxication with lesions of degeneration arising in a variety of tissues. These lesions of degeneration are to be seen both locally in the respiratory system as well as in distant parts, as in the muscles, kidney and liver. The primary damage arising in the respiratory organs, and which we believe to be the result of infection by the B. influenzÆ, facilitates attacks by such other bacteria as are available and pathogenic to man. The secondary invaders are not constant in type, but we find variations according to the localities where the epidemic takes place. Just as there is a difference in the bacterial flora which constitutes the secondary invasion, so, too, there is a variation in the picture of the inflammatory process which appears in the lungs. The occurrence of the miliary streptococcal broncho-pneumonia has been met with in certain localities much more frequently than in others; lobular and confluent pneumonia has been the prevailing type in certain regions, while a lobar purulent pneumonia with abscess and gangrene was most frequent with others. There does not appear to be an individual and constant character in the mode of distribution of the pneumonia in the lungs. That the pneumonias were not the usual type otherwise seen, is fairly agreed upon by all. The most astonishing feature presenting itself to us was the frequency of death occurring in the early stages of the inflammatory process and before the gray stage had definitely developed. The gray stage of influenza pneumonia is a purulent pneumonia which often also constitutes an acute interstitial pneumonia.

The extensive hemorrhage and inflammatory oedema of the lung are striking during the early stages of the lung involvement. The mononuclear infiltration which appears early and remains for a variable time, until the purulent process is well under way, is also unique. The hyaline deposit in the lung alveoli; the capillary thrombosis and necrosis of the alveolar walls and bronchi are important; while the tendency to abscess, infarct, gangrene and incomplete resolution with fibrosis differentiates this type of pneumonia from the common lobar variety.

As an organic evidence of the acute intoxication, none stands out more prominently than the degeneration of the voluntary muscles. These resemble the waxy degeneration of other bacterial intoxications, and particularly that of typhoid fever. The finding of these acute degenerations does not assist us in arriving at a conclusion as to the nature of the poisonous body, whether a true exotoxin. The presence, however, of such widespread degenerative lesions in cases showing no naked eye change suggests, at least, that the peculiar muscle weakness associated with pain has its origin in this definite process and not in primary nerve lesions.

Very interesting it is that the different muscular structures are not equally affected by the intoxication. This is particularly noteworthy in the heart and intestine. In neither of these structures have we met with lesions comparable to those in the voluntary muscles. Wherein this immunity resides we cannot state. In our own series, as well as in the majority of others, there was an unusual absence of evidence of myocardial weakness. In most of those dying during the acute illness, the heart muscle was found firm and the cavities not dilated. This finding was in striking contrast to that found in acute lobar pneumonia where dilatation of the right ventricle and auricle, along with muscle degeneration, is almost the rule. In but one case of the present series did we find myocardial degeneration leading to dilatation of the cavities and causing death. And in this particular case the intoxication was due to a streptococcus septicÆmia arising as a late sequel from the middle ear. The heart in influenza withstands remarkably well the effects of an intoxication from the disease and carries the extra load imposed upon it by the involved lung with little evidence of fatigue.

It is also worthy of attention to note that the kidney suffers so little in this severe disease. Bacterial localization with inflammatory concomitants does not occur, and there is no lasting damage upon its structure. As in so many conditions of bacterial poisoning, tubular degeneration, varying from a cloudy swelling to a more acute damage, is to be found in a percentage of cases, but complete restoration is rapidly obtained in convalescence. It is unusual to find such severe renal damage to incapacitate function to a degree to endanger life.

Finally we can add our evidence, gained from a study of the pathology of epidemic influenza, that the primary disease induced by the invasion of the B. influenzÆ opens the way for secondary infections of a variety of kinds, whose subsequent effect may be more serious than initial lesions. The many late complications which arise in this manner we have not investigated.

BIBLIOGRAPHY

Abrahams, Hallows and French	Lancet., 1919; i, p. 1.
Advisory Board to the D. G. M. S	Brit. Med. Jour., 1918; ii, p. 505.
Blanton and Irons	Jour. A. M. A., 1918; lxxi, p. 1988.
Boggs	Johns Hop. Bull., 1905; xvi, p. 288.
Brooks and Cecil	Brit. Med. Jour., 1918; ii, p. 496.
Chickering and Park	Jour. A. M. A., 1919; lxxii, p. 617.
Christian	Jour. A. M. A., 1918; lxxi, p. 1565.
Cole	Brit. Med. Jour., 1918; ii, p. 566.
Cole	Canadian Med. Assoc. Jour., 1919; ix, p. 41.
Dever, Boles and Case	Jour. A. M. A., 1919; lxxii, p. 265.
Fildes, Baker and Thompson	Lancet., 1918; ii, p. 697.
Fletcher	Lancet., 1919; i, p. 104.
Friedlander, McCord, Sladen and Wheeler	Jour. A. M. A., 1918; lxxi, p. 1652.
Goodpasture and Burnett	U. S. Naval Med. Bull., 1919; xiii, No. 1.
Hall, Stone and Simpson	Jour. A. M. A., 1918; lxxi, p. 1986.
Hunt	Lancet., 1918; ii, p. 419.
Keegan	Jour. A. M. A., 1918; lxxi, p. 1051.
Kuskow	Virchows Archiv., 1895; cxxxix, p. 406.
Le Count	Jour. A. M. A., 1919; lxxii, p. 650.
Lord	Boston Med. and Surg. Jour., 1905; cl, p. 537.
Lyon	Jour. A. M. A., 1919; lxxii, p. 924.
MacCallum	Jour. A. M. A., 1919; lxxii, p. 720.
MacCallum	Monog. Rock. Inst. for Med. Res., 1919; No. 10.
Muir and Wilson	Brit. Med. Jour., 1919; i, p. 3.
McMeans	Archives of Int. Med., 1917; xix, p. 709.
Nuzum, Pilot, Stangl and Bonar	Jour. A. M. A., 1918; lxxi, p. 1562.
Oertel	Canadian Med. Assoc. Jour., 1919; ix, p. 339.
Opie, Freeman, Blake, Small and Rivers	Jour. A. M. A., 1919; lxxii, p. 556.
Speares	Boston Med. and Surg. Jour., 1919; clxxx, p. 212.
Stone and Swift	Jour. A. M. A., 1919; lxxii, p. 487.
Symmers	Jour. A. M. A., 1918; lxxi, p. 1482.
Synnott and Clark	Jour. A. M. A., 1918; lxxi, p. 1816.
Torrey and Grosh	Amer. Jour. Med. Sci., 1919; clvii, p. 170.
Weber	British Med. Jour., 1919; i, p. 8.
Wittingham and Sims	Lancet., 1918; ii, p. 865.