glucose combines with two of diacetic acid. This is very analogous to what is supposed to take place in the animal body. Whenever there is a larger number of molecules of fatty acid being oxidized than of glucose in the metabolic mixture, then we would expect the excess of acetoacetic acid formed from the fat to be left unoxidized and on accumulation there would be ketosis.

In the effort to find out whether this idea is applicable in the human subject it was attempted to find out how many molecules of the so-called ketogenic substances are being metabolized in relation to the number of molecules, of glucose, when small amounts of the acetone bodies appear, or when there is a marked ketosis. In order to do this we must know not only how much fat, how much protein and carbohydrate is being burned at any one time, but we must have a means of estimating their ketogenic and antiketogenic influence in this reaction. I cannot take time to ask you to go into the calculations by which these factors, and the so-called "ketogenic balance" or or ratio are determined, but I would ask you to consider some of the deductions concerning ketosis in the body which follow from this point of view.

We have analyzed the ketogenic balance of a considerable number of subjects with varying degrees of ketosis-and some mild ones with only moderate ketosis, and a number of normal subjects with ketosis from fasting. In all of these it was observed that when ketosis first appeared or in the mild ketosis, the ratio between the number of molecules of the ketogenic substance to the num

ber of glucose molecules being oxidized at the same time, was practically constant, and that the ratio between them was about equivalent to one molecule of the total ketogenic substances to one molecule of glucose.

Further, it appeared that in those with marked ketosis, if one could estimate the number of molecules of fatty acid being oxidized and the number of glucose molecules being oxidized, that the excess of the former approximately represented the excess of the diacetic acid which failed to oxidize and was excreted in the urine. The conclusion drawn from these results seems to have certain necessary corollaries, which are of some aid in the attempt to formulate diabetic dietaries. The fuel burned in the body is made up of sugar, fat and the residues of protein. When, as in diabetes, the metabolism of glucose, whether derived from carbohydrates or the amino acids or glycerol of fats, is reduced by the low carbohydrate tolerance, the total fuel burned is not decreased. The diabetic keeps on metabolizing as much as the normal under the same conditions, and the result is that the amount, particularly of fat, which he must burn to keep up his total metabolism, when the carbohydrate is removed or largely decreased, becomes very great.

If it is necessary in order to avoid ketosis, to oxidize a minimum number of molecules of glucose, for a given amount of fat, as we have shown to be the case by the in vitro experiments, we can estimate about how much total glucose it is necessary for the patient to burn in relation to the height of his total metabolism, and thus can predict how much carbohydrate he

must burn in order to avoid ketosis. If his carbohydrate tolerance is not sufficient to permit him to burn that quantity of glucose, obviously he will have a preponderance of ketogenic molecules over the antiketogenic molecules and ketosis is inevitable.

By defining in this way the minimum amount of carbohydrate that must be fed as well as burned, one has a means of estimating the minimum dosage of insulin as well as carbohydrate, and it has been so employed in the standardization of insulin administration.

This point of view also, I think, makes clear the undoubtedly favorable effect of undernutrition in diabetes.

You will recall that previous to the discovery of insulin a favorite method of avoiding ketosis and coma in diabetes was by keeping the patient underweight, under-feeding him, and by so doing keeping him under normal metabolism. Not being able to push the carbohydrate tolerance up, as we now do by insulin, the other side of the ledger was pulled down and the amount of ketogenic fat and protein burned thus kept within the limits imposed by the low (antiketogenic) glucose tolerance. Many have had the experience of having a diabetic patient go along pretty smoothly and then develop a tonsillitis or some other infection, with the sudden appearance of ketosis and coma. That is un

doubtedly due to the effect of the infection speeding up the total metabolism, with a consequent increase in the metabolism of ketogenic fat and amino acids and perhaps a decrease of carbohydrate tolerance, resulting in a marked disturbance of the ketogenic balance. Under such conditions serious ketosis may be inevitable unless the carbohydrate tolerance is maintained by insulin.

This point of view I think allows us to see the simple qualitative tests of the urine in a clearer light. Obviously, a glycosuria without a ferric chlorid reaction for diacetic acid shows that the carbohydrate tolerance is lower than the carbohydrate ingested but is not so low as to provide a negative ketogenic balance. If there is both glycosuria and ketosis, both sugar and diacetic acid, you know that the carbohydrate tolerance is so low that, at the rate of that subject's total metabolism-the rate at which the subject is living a ketogenic balance cannot be maintained, and anything that disturbs it further may produce a marked acidosis and coma. ger line is crossed. The patient should be in bed and probably will need insulin. If on the other hand there is ketosis without glycosuria it probably means merely that the glucose or carbohydrate supply is insufficient to provide a ketogenic balance and that that is not due to a lowered carbohydrate tolerance.

The dan

though, by this method have been rather disappointing.

That leaves only a few cases which we feel require major surgery, meaning by that drainage or anything which may follow along after the institution of drainage. Unfortunately, this constitutes the most distressing type of abscess. As you have all doubtless had the experience, a large abscess with a large amount of pus will respond very readily to drainage. The patient will improve readily, the cough will disappear, he will rapidly gain weight and so on, but unfortunately he often will not get entirely well and will continue to have a cough with expectoration, occasionally a hemorrhage, clubbing of his fingers and his toes will become progressive and he will be in a state of indescribable misery and will perhaps be an outcast from society. This condition does not necessarily mean that he has developed a bronchiectasis in the ordinary sense of the word. He may have a lung honeycombed with small abscesses without much bronchiectasis. On the other hand, he may have a frank bronchiectasis with greatly dilated and thickened bronchioles.

Now I think we have made considerable advance in the possibilities of treating these miserable people who

have these chronic suppurations and it is that type of case particularly which I wish to take a few minutes to deal with this afternoon. Perhaps I can do this best by means of some lantern slides. (Then exhibited series of slides.)

In the chronic conditions in the most marked type, namely, bronchiectasis, we find a condition with which you are all familiar. This drawing (indicating slide) represents the lower lobe of a man who had bronchiectasis for fifteen years. He was a social outcast and had been in tuberculosis sanitaria for four years, but never had tuberculosis. It is quite evident that simple drainage would accomplish nothing here. In all of these chronic suppurative conditions the ideal thing is to get rid of that diseased tissue just as we get rid of the chronically inflamed appendix or gall-bladder. Other measures which compromise on that idea will frequently give temporary but very rarely permanent relief. One method is the thoracoplasty used twenty-five years ago by Tuffier and others and discarded. It has recently been revived by Hedblom of Rochester, who has reported some excellent results. We think the ideal thing is to get this tissue out and be rid of it. The old operation of lobectomy was the most satisfactory theoretically, as far as getting out the diseased tissue was concerned, but it has a reported mortality of 52 per cent and the patients which survived the operation are not all well. We therefore felt that the operation of lobectomy as ordinarily carried out carried a prohibitive mortality and we did not feel justified in recommending it. Before devising our new opera

tion, however, we performed the old operation in a few cases and this (indicating slide) shows the final result in a case with complete healing, three years afterward. The patient was completely restored to health and is now an amateur pugilist.

This diagram (indicating) shows why it is very difficult to carry out lobectomy. We must first free the lung of adhesions. The lung is nearly always adherent to the chest wall, the pericardium and the diaphragm. We then nearly always have exposed some of the diaphragm, the mediastinum and the pleura to an infection which seems to be inevitable. The serious thing about the operation is the inevitable infection of the raw tissue exposed. Most of the deaths have been due to a suppurative mediastinitis.

In order to accomplish the removal of the chronically diseased tissue in the lung we devised a new operation which we have performed in 17 cases, with 1 operative death.

This picture (indicating slide) represents an old, chronic lung abscess which had been drained but not made well. The drainage tube runs right up to the hilum.

The operation consists in, after making a free exposure of the diseased tissue, going in with the actual cautery, a soldering iron, and burning out all all the diseased lung tissue. The advantage of that is that it is not necessary to mobilize the part of the lung that has to be removed, it is not necessary to free it from adhesions and, therefore, it is not necessary to expose a portion of the mediastinum and surrounding areas to infection. All this portion (indicating) was burned out with the actual cautery and you can see the results three weeks later here (indicating). This was taken with a little rubber bag with barium in it to show the amount of diseased tissue removed. This patient is now entirely well except for a small bronchial fistula and she has been restored to former activity and is entirely free from cough and expectoration.

With the exception of one sudden death, which occurred probably from embolism of the brain, we have had no untoward effects. One other patient, however, died of pneumonia in the other lung six weeks after the operation.