produced they will be the same distance as in the original instance. The positive thus produced will be an outline of the superior strait in its true dimensions and diameters may be measured directly with a centimeter rule.

It may be useful to mention certain points in technique that have helped in developing the process. In order to secure accurate focusing it is useful to outline on the negative by means of a red wax pencil or india ink the outline of the superior strait and the ends of the pelvimeter. This produces in the positive a white line. from which the various diameters are easily measured.

In obese subjects in order to shorten the necessary exposure required the target is best used a little nearer the plate, say three feet. The distortion thus produced is of course somewhat greater but is easily corrected when making the positive picture.

A word or two might be said as to the applicability of the method to pregnant patients. There is no reason why it should not be used during the first six months of pregnancy. Later than this the amniotic fluid and increased size of the uterus and fetus would probably form obstacles which would not permit of good pictures.

The question has been raised from time to time as to the possible deleterious effect of the roentgen ray upon the fetus in utero. Edelberg in 1914 showed that the danger of injury by this means was negligible. He observed the condition of a child which had been conceived while the mother was under the effect of full X-ray treatment, one quarter of the total quantity of the treatment falling into the period of gestation. The child was born at term fully developed with all signs of maturity. Nothing pathological was noticed and the child was well nourished.

In the early developments of this method numerous experiments were made photographing dried pelves. It was found that a positive image could be produced which would be accurate almost to a millimeter. Later the direct conjugate was measured on a patient during a laparotomy. Following her convalescence she was measured by means of the above method and a pelvigram of her pelvis was produced which was accurate in its antero-posterior diameter almost to exactness.

In conclusion I wish to thank Dr. Louis H. Wheatley, of New Haven, the roentgenologist at Grace Hospital, for his interest and co-operation in the preparation of this work. Not only has he been most helpful in the preparation of the pictures but to him belongs the credit for working out the exposures and other points of roentgenologic technique.

59 College Street.

DISCUSSION.

DR. LOUIS F. WHEATLEY, New Haven: My part in this work has been of a very minor character. In the ordinary semirecumbent position, at six foot distance, we were getting very clear plates of the pelvis, but there was an absence of the bulging promontory of the sacrum which showed that we were not getting a true perpendicular. After various methods were tried, we finally obtained this by causing the patient to assume an exaggerated arched position of the back, so that we could direct the central ray through the median line one inch below the umbilicus. This gave us the desired outline of the promontory of the sacrum, and a true perpendicular of the pelvic inlet. With the patient in this position on the Bucky diaphragm, the pelvimeter points were applied to the hollow of the sacrum and symphysis pubis and the arms of the pelvimeter held parallel to the plate during the exposure, thus insuring a true horizontal position of the points to the plate. We used a seven and one half spark gap, twenty milliamperes of current, twenty to thirty seconds of exposure. The distance, originally set for six feet to minimize distortion, was reduced to three and one-half feet, because of too much strain on the tube. Even at this distance, we had to interrupt the exposures because of overheating. The greater divergence of rays incident to the shorter distance was overcome by a proportionate reduction of the negative to the positive image.

I believe that, in this manner, we are able to obtain very accurate dimensions of the pelvis. We have thus a true perpendicular, a true horizontal, and the ratio between the external conjugate as measured by the pelvimeter points on the skin surface and that measurement arrived at by the proportionate reduction between the pelvimeter points and distance between the symphysis and promontory of the sacrum as measured on the positive from the X-ray film.

I feel that this should be of considerable value in questions of contracted or deformed pelves and border line cases.

It is a very simple procedure and does not call for any complicated geometric equations and formulae as several other methods do, and it can be carried out by any one with suitable apparatus and a simple knowledge of proportion.

Dr. Thoms is certainly to be congratulated on developing such a simple method of such value.

DR. JAMES R. MILLER, Hartford: The Society is to be congratulated at having presented at this meeting such a clear demonstration of a new diagnostic procedure. Dr. Thoms has combined well known facts and taken advantage of processes which were at the disposal of every one, in a very ingenious way.

The method commends itself because of its accuracy and simplicity.

I would like to add a word to his statement about the harmlessness of the procedure. There are a great many reports in the literature at present which show that even after a very thorough X-ray treatment, such as is given for fibroids, pregnancies may occur and that the products of conception are in no wise interfered with.

Paul Werner has reported seventeen such cases with a total of twenty-four pregnancies observed over periods up to eight years. He believes that there is very slight increase in the tendencies to abortion and that the children are very slightly under weight as they develop years afterwards.

It is to be remembered, however, that all of these cases received intensive X-ray treatment with many times the quantity of rays which are necessary for the diagnostic observations of the method under discussion.

Since reading Dr. Thoms' paper, three days ago, I have been able in two cases to check up the internal measurements by this method with the aid of the Gauss pelvimeter, which I believe to be the most accurate instrument of its kind. The only difficulties I can see are purely technical ones which experience will easily correct. The measurements are difficult in a very fat woman, but are quite accurate in a thin woman. This method is, of course, limited in its application from a practical standpoint, that is, it will be seldom necessary to use it. There are cases, however, where both the doctor and the patient will be very glad of the graphic and accurate measurements of the most important pelvis diameters. I refer particularly to cases of congenital hip disease, especially in nullipara and in conditions which are apt to distort pelvic bones during the developmental period in an irregular

manner.

With this method at our disposal, we may be able to assure ourselves and the patient in many cases that there will be no pelvic dystocia provided the child is not too large.

I should like to congratulate Dr. Thoms on his extremely ingenious method.

Observations on the Source of Infection

in Pneumonia.*

DR. FRANCIS G. BLAKE, New Haven.

Progress in measures for the control and prevention of infectious disease demands an accurate knowledge of the source of infection, of the mode of transmission of the infectious agent from one individual to another, and of the factors that determine whether disease will result when transfer of the infectious agent has taken place. As familiar examples of the results obtained following the acquisition of such knowledge one may cite typhoid fever, bacillary dysentery, yellow fever, malaria, and diphtheria. In spite of the brilliant achievements of preventive medicine in the control of these and other diseases, comparatively little has been accomplished in the field of acute respiratory infections of which pneumonia is an outstanding example. It has seemed, therefore, that it might be of some interest to examine such knowledge as has been gained during recent years concerning the source of infection in pneumonia with the purpose of raising the question whether we are utilizing this knowledge to the best advantage in an effort to lessen the continued high incidence of this disease.

Theoretically two points of view may be considered concerning the source of infection in pneumonia: (1) that pneumonia develops in an individual who normally harbors the pneumococcus, hemolytic streptococcus, influenza bacillus or other causative organisms in his upper respiratory tract when contributing factors affecting the normal resistance of the individual make it possible for otherwise harmless bacteria to invade the lower respiratory tract and produce pneumonia. In brief, this is the theory of autogenous infection. (2) That pneumonia is due to an invasion of virulent bacteria from outside the body, that is to contact infection, either direct or indirect, from exogenous sources. In this case it is conceivable that pneumonia might inevitably arise whenever

*From the Department of Internal Medicine, Yale University School of Medicine.

transmission of the organism in question takes place, or on the other hand that it might arise only when further contributing etiological factors affecting the individual become operative.

It is obviously of the greatest importance to know which of these two possibilities is the correct one from the point of view of practical prophylaxis. If infection is autogenous in source little can be accomplished by quarantine measures and efforts would of necessity have to be directed entirely toward the prevention of the contributing etiological factors. On the other hand if infection is exogenous in source, measures directed toward the prevention of the transfer of the infectious agent from individual to individual become of great moment, while elimination of contributing factors may or may not be of importance depending upon whether the development of pneumonia is or is not inevitable in the presence of the organisms causing it.

For the purpose of the discussion in hand it seems desirable to divide pneumonia into two groups, (1) lobar pneumonia, a specific acute, infectious disease caused by the pneumococcus; (2) all other cases of pneumonia, usually secondary to some preceding infection and caused by a variety of bacteria, the more important of which are pneumococcus, Streptococcus hemolyticus, Bacillus influenzae, and Staphylococcus aureus. This group usually presents the clinical picture of bronchopneumonia; it may, however present that of lobar pneumonia.

Lobar Pneumonia.

For many years following Frankel's demonstration in 1884 that the pneumococcus was the bacterial cause of lobar pneumonia, it was held that the presence of the pneumococcus was only of minor importance in a series of phenomena leading up to the onset of pneumonia. This view was inevitable as long as all pneumococci were held to be identical, since it had already been shown by Pasteur in 1881 and independently by Sternberg in this country at the same time, that pneumococci virulent for animals were frequently present in the mouths of normal individuals. In other words it was generally believed that infection is autogenous,--that is the individual becomes infected with the organism already