are many kind-hearted officers; there are numbers of excellent diet lists filled up for these men to tempt their ever-ebbing appetites; but still the consumptive sailor dies just as soon and just as surely under one man's treatment as under another's. It is a sad fact that we are not curing these cases and our results are bad, though we probably get as good or better results than civil hospitals in the same climates.

WHY THE RESULTS ARE BAD.

There can be no doubt that the results are very bad, and principally from three causes: (1) Overcrowding; (2) the poorest of climatic conditions at all of our stations where there are hospitals; and (3) from the fact that the disease is rarely diagnosticated in the first stage, the only time when treatment might prove efficacious even in a poor climate.

OVERCROWDING.

When it becomes necessary to put consumptives in wards with patients suffering with other diseases then a serious error is being committed, and the defect of a hospital which makes such a condition necessary should be quickly remedied. That such conditions do arise in many of our hospitals is unfortunately true. It is also true that consumptives have been put in with other patients to save the expense of heating, lighting, and administration of large wards, when there were only a few consumptives on hand. Putting consumptives with other patients has been officially prohibited by the Surgeon-General.

Many of our hospitals have been built on the pavilion plan or with long wards. They were constructed without regard to isolation of any class of diseases, and so it is that-using Chicago as an examplethese cases, comparatively few in number, can not be segregated, except in very large wards, which are too large for economical administration. Similar conditions exist in other hospitals, and thus it becomes a temptation, when there is lack of room for other patients, to economize and put the consumptives in general wards, but this should not be done.

EFFECT OF CLIMATE.

One cause being pointed out for the great mortality of these patients, it may be of interest to examine the statistics in regard to the outcome of the disease as treated in hospitals in different parts of the country, showing effect of climate on the disease.

It has long been known and granted that mere equability of climate is of little, if any, benefit to the consumptive. There are few climates at the sea level that are not subject to rapid changes in temperature and humidity. None of our hospitals are situated in dry climates, though several stations have heat enough, and all are humid and saturated. It is probably true that a moderately cold and dry atmosphere at high altitudes, without wind and great temperature changes, is the best for the consumptive. Such a climate is hard to find, and for many reasons would not be attractive as a subject for this paper. The next

best climate is the dry, warm atmosphere in moderately high altitudes, and is the one best suited for a marine-hospital sanitarium. The very worst of all climates is the hot, humid, saturated, devitalizing atmos phere at the sea level, which saps the strength and life of these patients in so short a time. We have no station where there is a hospital that would in any way answer the necessary climatic requirements for the consumptive sailor.

It will now be interesting to study the following tabulation of figures, and it will be well to bear in mind that it is necessary to avoid several sources of error in compiling such tables from the annual reports.

MORTUARY TABLES.

One of the greatest errors that will readily occur to officers is the fact that one person suffering with consumption becomes during the fiscal year probably two or three cases of consumption. These men come and go, being reported as new cases each time they enter the hospital, thus lowering the percentage of mortality by increasing the number of cases.

A study of the mortuary tables of the service for a number of years shows that the mortality is as great at the sea level in one district as in another, regardless of equability of climate. The percentage of deaths from 1880 to 1895 is 22.4, and from 1873 to 1877, 21.5.

Percentage of mortality among cases of tuberculosis from 1880 to 1895, inclusive, arranged by districts and in periods of five years:

EARLY DIAGNOSIS.

The early diagnosis of the disease in the seafaring class is a subject the great importance of which, I believe, does not sufficiently impress us in the Service. Too many of these cases are overlooked in the haste of dispensary practice, and obscure cases with coughs are allowed. to run on and on until it is possible and imperative to make the diagnosis by means of the microscope-the diagnosis seeking us, being forced on us, in fact. Every cough that has shown any disposition to continue after ten days' treatment should be looked upon with suspicion. All patients with coughs deserve thorough chest examination, and if there is expectoration, then use the microscope. If the bacillus is not found, do not be too sure even then that your patient is free from infection. If a chronic and a persistent cough proves not to be due to tuberculosis, it is more often found to be caused by granular pharyngitis rather than "chronic bronchitis."

We ought to be exceedingly careful in dealing with chronic coughs, for I fear that we are treating many cases as bronchitis that should be classed and diagnosticated as consumption. Turning to the annual report of 1895, as an average year, we find that 3,308 cases of acute, 558 chronic, and 81 cases of catarrhal bronchitis were treated. Of this number, 92 cases of chronic bronchitis were treated in hospitals, 28 of which are reported recovered (30.4 per cent), 46 are reported improved (50 per cent), 8 not improved, and 4 died. A guess that over 50 per cent of the chronic bronchitis cases were tuberculosis pure and simple would probably not be wide of the mark.

A thorough examination ought to mean that the patient be divested of all clothing covering the chest, and the thorax gone over inch by inch. I can not but look with suspicion upon the diagnosis (or the nondiagnosis, for that matter) of the man who uses his ear only, and that separated from the chest wall by clothing. When the physical signs of consumption can be easily detected through clothing by the unaided ear, then the disease has become seriously advanced and can not be checked except by heroic treatment.

Writers refer to the first stage of the disease, meaning that period. from infection to the date at which the bacillus can be found in the sputum. There are certainly many physical phenomena during the development and transition of this stage to aid one in making a fairly accurate diagnosis even before the baccillus can be found and before there is expectoration. That the disease is diagnosticated in this stage I believe to be rare.

If we could make a diagnosis before the bacillus has broken free into the air spaces of the lungs, and is found freely in the sputum, much good would result to the patient, and also probably prevent him from infecting others.

There are many symptoms which help the observing physician in making a diagnosis in this stage, and I trust that I may be pardoned

for pointing out and emphasizing a few which have been of most value to me.

The general appearance of most of these patients is characteristic and peculiar, and if one notes closely the facies many cases can be pointed out in that manner alone. The red lines on the gums, as pointed out many years ago by Frederieq and Thompson, is a very valuable sign, but it has not always been present in cases under my observation. There is usually a change in the timbre and pitch of the voice, even when there is no cough. Loss of weight usually comes early, the chest wall being the first part to show emaciation. In this stage percussion does not help one much, for it is next to impossible to detect a flattening of the percussion note; but in auscultation we have one of the surest aids to diagnosis-that is, to this date, and until the radiograph is further perfected.

That a diagnosis can now be made in the early stage by the radiograph is undoubtedly a fact. Much good work is now being done in making out cases in which physical signs are too obscure to be made out by ordinary diagnostic methods.

If one will bear in mind to seek for the first and initial lesion in the pleura, it will be surprising how often the disease can be first located as a small area of dry pleuritis. I believe that a large majority of tuberculous infections first involve the pleura, and that the first manifestation of the pathological condition is a small area of apical pleuritis without effusion, eventually resulting in agglutination of the two pleural surfaces and subsequent infection of the lung at the point of contact.

The following quotations from Whitney, Twentieth Century Practice, vol. 7, page 11, will show some figures collected to point out how frequently tuberculosis follows serofibrinous pleuritis:

If it can be shown that a considerable proportion of those who have had one or more attacks of serofibrinous pleurisy subsequently develop a distinct form of tuberculosis, the evidence thus afforded must be allowed to be of incontestable value. It must also be granted, as is so strongly insisted upon by Netter, that negative statistics have but little worth as against even a small amount of positive evidence; there are so many possible sources of error when, for example, it is stated that of a given series of effusions nearly all of the persons have continued in perfect health. The following are some of the most careful observations as to the subsequent history of pleuritics: Handford reports 5 cases of effusion in apparently well subjects, 3 of whom afterwards became tuberculous. Fields had 112 cases, of which number 25 subsequently died of tuberculosis and 65 became tuberculous. Barrs ascertained in 1890 the condition of 57 cases treated at the infirmary of Leeds from 1880 to 1885; 21 had died of tuberculosis. Bowditch investigated in 1889 the subsequent history of 90 patients, treated for effusion by the elder Bowditch from 1849 to 1879; 30 had died of tuberculosis, and at least 1 of those still living was tuberculous. Recochon was able to follow carefully 32 cases of pleurisy in private practice; tuberculosis developed sooner or later in all but 1.

Many similar series have been reported. Those selected, while showing a far greater percentage of tuberculosis than many others, are at the same time the most trustworthy at command. They represent cases seen in both hospital and private

practice. Taken together they give a total of 310 cases of pleurisy with 178 instances of subsequent tuberculosis, or 57 per cent. This is probably above the truth.*

Probably all pleurisies without effusion are due solely to infection with the tubercle bacillus. Necropsies on old tuberculous cases, when there is extensive pleuritic inflammation, rarely show effusion.

The portion of the pleura involved is usually that part above the third rib, and the patient can usually designate a point which is painful on pressure. On listening over the infected area, have the patient take a very deep inspiration, and hold the breath as long as possible. He will be unable to entirely suspend expiration, and one will then hear if there is pleuritis, without adhesion or effusion, a far-away roaring sound, something like the hum of machinery or the "waterfall" sound. The true pleuritic friction sound is altogether different. A very full inspiration and gradual expiration also produces a muffled sound in health, and must not be confused with the other. The early case also has a jerky expiration. To obtain the sounds perfectly the lungs must be fully distended, and the sounds referred to are produced by the inflamed and roughened pleural surfaces slowly and continually sliding over each other during gradual expiration. If there has been agglutination of the two pleural surfaces or there is a small sacculated pleurisy or lymph deposit one will not hear the waterfall sound in gradual expiration over that part. I rely a great deal on this symptom in making a diagnosis in the early and obscure cases.

Of course all are familiar with the sounds caused by destruction of the alveoli, due to perialveolar deposits of tubercles. Here the harsh, blowing sounds, or the gurgling of mucus, with absence of vesicular breathing over that spot, is itself of enough importance to justify a confirmatory diagnosis.

Another great help in ferreting out the disease in the early stage is the continuous temperature chart. I would advise against relying on the morning and evening reckonings alone, for while the chart made up this way will be valuable it will often be more helpful if the notings be made between 1 and 4 o'clock in the afternoon, or sometimes between 11 and 3 o'clock. If the daily curve shows a constant subnormal or a constant rise of temperature the case is very suspicious, and must be immediately differentiated from the few other diseases that might cause the notings on the temperature chart.

Examination of the blood will be of great value, though not determinative, as the abnormal conditions found at this stage are noted in other diseases. The percentage of hæmoglobin should be taken and a blood count made as matters of routine. If the percentage of hæmoglobin is quite low, the count low, and there is a leucocytosis, then we could exclude tubercle of the lungs in the early stages before it has become possible for a mixed infection. Holmes reports in a series of examinations that there is no leucocytosis in the early stage, and Cabot says there is no leucocytosis until there is a mixed infection or just 10918-16