Roentgenological methods clearly demonstrate that tuberculosis may begin around the roots of the lungs, where physical examination fails to give satisfactory results, much more frequently than had previously been supposed.

According to Tendeloo primary foci occur chiefly in the peribronchial and perivascular tissue; and it is possible that bacilli either pass directly from the pulmonary lymphatics or do so after they have infected one or more of the peribronchial lymphatic glands. If this be acknowledged, probably the lympathic tissue and the glands about the hilus become affected quite early, even before parenchymatous changes are demonstrable.

Opie holds the opinion that approximately 50 per cent of adults have encapsulated lesions in the lungs or the bronchial lymphatic nodes. Discussing the question of tracheo-bronchial glands Tendeloo quotes Nageli's figures showing that among III cases of latent, inactive tuberculosis of the lungs and glands, the tracheo-bronchial glands alone were diseased in 16 cases. It is surprising to what an extent the hilus region is involved in patients with acute respiratory infections. The glands are enlarged and there is often pronounced congestion of the surrounding tissue which may not subside until after the ordinary physical signs have disappeared. In chronic cases, even with frank tuberculosis, the glands are often enlarged and present many small calcified areas. In both acute and chronic cases where parenchymatous changes in the lungs are demonstrable only with difficulty, or not at all, the peribronchial tissues and at times the pleura also are found thickened.

In 1911 in a study of the mediastinal glands, I expressed the conviction that enlarged glands may be responsible for the signs interpreted to mean an early apical tuberculosis, and furthermore that they are as common in adults as in children. At that time, however, I did not appreciate that enlarged glands often represent the starting point for later pulmonary tuberculosis. In this sense they thus afford opportunity for a very early diagnosis, the value of which is widely appreciated and has been especially emphasized by Overend and Hebert. These writers also state that the types of pulmonary tuberculosis most liable to render the clinical diagnosis uncertain are:

(a) A purely glandular type, affecting the hilus, bifurcation and paratracheobronchial lymphatic glands, separately or collectively.

(b) A simple or attenuated form of peribronchial tuberculosis.

(c) The peribronchial disseminated type.

(d) The central or hilus type, in which the parenchyma of the lung becomes gradually infiltrated from the hilus region.

With this classification I agree for in a long series of cases I have observed all these varieties. Moreover, my experience convinces me that the first and fourth types are particularly important. I shall later demonstrate these types by means of lantern slides.

Adenopathy of the hilus group, as has been pointed out, may be directly related to the condition of the lung-a fact upon. which Barjon lays particular stress for he is convinced that the development of pulmonary tuberculosis is often attributable to the gland lesions. My own opinion is that any involvement of the hilus region, whether lymphatic, pleural, glandular or otherwise, may indicate either a pre-pulmonary tuberculosis or an early pulmonary lesion. The enlarged glands alone may cause no special symptoms. They are found, as I have said, in other acute and chronic conditions less serious than tuberculosis; on the other hand, when associated with other clinical signs, as rise in temperature or changes in breathing, they become of great diagnostic significance.

Now, radiologically the hilus glands give a fairly definite image, though at times they may be confused with adjacent structures. Typically the hilus is separated from the median shadow, but, if the adenopathy progresses even a little, the hilus and the median shadow fuse. Again, the hilus shadow should be nearly homogeneous. It should be fairly clearly outlined, and somewhat crescentic in shape. On the other hand, if glands or inflammation of the hilus tissues are present the shadow is enlarged and becomes less homogeneous. On the left the hilus shadow is rarely well seen, for it is covered more or less extensively by the heart.

In the presence of an extensive and diffuse shadow extending

from the hilus region into the lung probably pulmonary involvement exists. Clinically, tracheo-bronchial adenopathy is associated with mediastinal adenopathy while radiologically they are separate. Barjon believes that this distinction is artificial, as mediastinal adenopathy in children is usually accompanied by a hilus adenopathy; in adults, hilus adenopathy usually ends with mediastinal adenopathy.

My observations in tuberculous cases indicate that not infrequently at a given moment the greater the hilus involvement, the less the affection of parenchyma and vice versa. In acute diseases this rule is reliable in the majority of cases, for example, in frank and broncho-pneumonia. In these cases it serves as an indication of the extent and acuteness of the process. Moreover, in certain cases of tuberculosis, especially of the peribronchial type, the shadow cast by the hilus region enlarges progressively to a certain point and then contracts, coincidently becoming denser. With tuberculosis extending from the hilus, the shadow cast by the region undergoes progressive changes similar to those in the parenchyma until the process in the lungs. is extensive and then it may be termed a pulmonary form. From this time, clearly, the hilus changes are unimportant.

In a number of cases I have found that before definite, well defined clinical signs of tuberculosis were present, the hilus region was reacting to the infection; and later the parenchyma became affected as the hilus signs were subsiding. Heretofore, the changes at the hilus in so-called pre-tuberculous cases have been considered irrelevant. In my own experience it was not until several cases returning for re-examination through a period of from three to six months or longer showed parenchymatous changes with disappearing hilus changes that my attention was directed to this region and the sequence of phenomena it presented.

The hilus region then, including the lymphatics and the glands, possesses no little clinical and radiological significance—a problem upon which I am working at present. While this statement of my findings is intended as a preliminary report, it will serve to draw attention to the fact that gland enlargement and

hilus changes must be taken seriously into account in the interpretation of radiographs for the diagnosis of early pulmonary tuberculosis.

DISCUSSION.

DR. BLUMER (New Haven): Mr. Chairman, and members of the Society. Before the advent of the X-ray in the diagnosis of disease we were dependent for our interpretation of disease processes on what we could observe clinically and what we could observe on the post-mortem table, and it is obvious that there were very serious drawbacks, especially in certain situations, in connection with the clinical and pathological points of view.

So far as the particular region under discussion is concerned, you all ought to know through practical experience that it is extremely difficult to diagnose, particularly in adults, pathological conditions in connection with the hila of the lungs; much more difficult than it is to diagnose changes in other portions of the lungs which are more accessible. In connection with the pathological observations it is well to bear in mind that in drawing conclusions from what we find on the autopsy table we are always laboring under one great disadvantage, and that is we only see the end result. Although we may have a series of cases we are frequently unable to make out a definite chain of pathological circumstances from the beginning to the end of the disease, and it seems to me that one of the important things that this paper brings out, is that the use of the X-ray may serve as an important link-I might also almost say a missing link-between the interpretation of the autopsy room and the interpretation of the bedside. It is likely to serve as an extremely important link between these two kinds of observation.

Of course, the paper brings up an old question that has been discussed a great many times before, and that is the relationship which exists between tuberculosis of the bronchial lymphatic glands and the tuberculosis of the parenchyma of the lungs. It seems to me that the point that Dr. Honeij has brought out throws a great deal of light so far as adults are concerned. For a great many years there has not been very much doubt about the relation between tuberculosis of the bronchial and the mediastinal glands in infancy and childhood and tuberculosis of the lungs in infancy and childhood. I think the reason that there has been no doubt in those cases is that tuberculosis in infancy and childhood is a progressive disease. We do not see tuberculosis regress in infants and children. At any rate, we don't see pulmonary or thoracic gland tuberculosis regress in children to anything like the extent we see it in adults. So on the autopsy table in children, even though the patient has lived for a considerable length of time, we find definite evidence both in the

glands themselves and the lungs that indicate the relationship between the two processes.

Now, as Dr. Honeij's paper brings out particularly, the thing that has misled us in our interpretation in the adult of the relationship between tuberculosis of the bronchial glands and tuberculosis of the lungs is that in a certain proportion of cases of pulmonary tuberculosis of the adult there is apparently a primary disease in bronchial lymph glands, and then a secondary infection of the lungs, and as the pulmonary process progresses the gland process regresses. You can readily see what that means from the pathological point of view. It means that when the patient has finally succumbed to the pulmonary process the glandular process has regressed to such an extent that we have been led to a probably false conclusion that there is no definite relation between the glandular process and the pulmonary process.

It seems to me that one of the chief points of interest that this paper illustrates is that one group of cases-we don't know quite how large a group of pulmonary tuberculosis in adult life, is similar to those cases which occur in childhood in which the primary lesion is in the bronchial lymph glands and the secondary lesion in the lungs.

The only other point which occurs to me is whether the X-ray may not serve, as the electrocardiograph has served, as a method of putting us on the track of physical signs which we have hitherto overlooked. The use of the electrocardiograph was instanced because it is such a striking illustration. You all know that not many years ago when we recognized an irregular pulse we were not able to determine what type of irregularity it was. With the use of the electrocardiograph and instruments of precision we have now reached the point where we can detect with our ordinary unaided senses simply by feeling the pulse many of the different kinds of irregularity.

While this is a very difficult region to explore from the point of view of physical diagnosis, while the physical signs that have been described as indicating involvement of the hilum region are comparatively few, it occurs to me that there is a possibility that by the use of the X-ray in conjunction with methods of physical diagnosis we may be able to work out clinical signs that are much more accurate than those we possess at present,

DR. OSBORNE (New Haven): Mr. President, I don't believe I have any more to add. I think, as Dr. Blumer has already stated, it is very fine that we can get positive testimony of what these glands are doing and can do. Those of us who see the clinical side of things constantly realize how many, many times incipient tuberculosis gets by us even in the adult, and we feel that such testimony as this is one added thing to help us make our decision. Of course, there is another side, that we must not