Even when several kinds of blood are mixed, and in high dilution, the specific reaction for each one can be obtained. Thus Nuttall after mixing five different bloods together so that each was diluted about 1:500 was able to get the specific reaction for each with its anti-serum, but no reaction with anti-sera for other kinds of blood than those present in the mixture. Stern increased the precipitin in one rabbit's serum so that when added to human blood diluted 1 : 50.000 it gave a positive reaction.

That more work is required to determine the precise conditions under which this test is applicable and its exact limitations is evident. But, on the other hand, in those important cases of not uncommon occurrence where the ends of justice require an accurate differentiation between certain kinds of blood stains, this method certainly promises more help than any other yet known, and we may expect will supplant the attempts to obtain this differentiation by the determination of the size of the corpuscles.

Instances of the application of this serum method are already at hand. Here in New England I have noticed brief reports of two cases in which it has been used, and with evident satisfaction. One of these was by Dr. Whittier, of Maine, and the other by Dr. Wood, of Boston. As emphasized by Dr. Whitney, in any case of death by violence where "there is a possibility of a blood stain having to be examined, a strip of filter paper should be soaked in the blood of the individual at the autopsy." This should be dried and preserved and later can be tested with anti-serum in order to set at rest any doubt about that individual's blood giving the reaction.

RELATIVE NUMBER OF CONTRACTED PELVIS IN

GENERAL PRACTICE.

PATRICK CASSIDY, M.D.,

NORWICH.

The general practitioner, although having a limited number of obstetrical cases in the course of a year, meets with many cases of difficult and prolonged labor. These cases are due often to causes which may easily be accounted for without any mechanical aid and may be classified as, 1st, those due to advanced age in primipara; 2nd, those due to adverse presentations, as breech, or face, or transverse, etc., in women having good pelves; 3rd, the so-called dry labors; 4th, those with insufficient or irregular pains; 5th, cases of very large child. But there are a certain number of cases in which everything seems to be favorable to an easy labor but in which labor is prolonged and difficult, sometimes requiring interference by the obstetrician and in which there is a great deal of mental and moral perturbation on the part of the patient and her friends. These in many cases are due to a slight contraction of the pelvis. Partly to be abie to foresee these cases and so prepare the minds of the families and partly from personal interest, I have during the past year made pelvic measurements on my new patients and such of my old as came under the class last mentioned. I have kept records of thirty consecutive cases and will present my findings in this paper.

The question of the relative number of occurrences of contracted pelvis has been considered by many observers in this country and abroad during the past ten years. The observations in this country have caused us to recede from the attitude taken, even so late as 1896, that

among the American-born women the occurrence of contracted pelvis was very rare, and was to be found principally among women born on the continent of Europe. Lusk, in his text-book on midwifery, makes the statement that, although among the women of foreign birth, confined at the Emergency and Maternity Hospitals, the average frequency of contracted pelves was fully up to the fourteen per cent. of Michaelis, Spiegelberg and Schroeder; yet among the native American women the condition was rare. He adds, however, that he believes the condition in many cases to be overlooked. J. Whitridge Williams published in the Bulletin of Johns Hopkins Hospital in 1896 the results of examination of one hundred pelvic measurements taken by him, and found in this number fifteen cases of contraction. In eleven of these cases the women were of American birth. In this paper our first consideration is as to what we shall consider a contracted pelvis. If we should call all deviations, on the small side, of the external pelvic measurements to represent pelvic contraction, then out of the comparatively small number of cases here considered I found thirty-three per cent. of contractions. But considering as contracted only those pelves that have a deviation in transverse measurements of more than three centimeters, and of antero-posterior diameters, two centimeters contraction is found in ten per cent. of the cases measured. Of these, one was a generally contracted rachitic pelvis with more or less obliquity due to Scoliosis. In this case the external conjugate measured eighteen centimeters and the oblique conjugate between seven and eight centimeters. Caesarian section was advised, but the patient refused to consider it, and after attempted delivery by version, Craniotomy was of neces sity performed on the dead child. In the other cases of contraction the pelves were of the justo-minor type. one, the presentation was occiput posterior, and, after a difficult long forceps operation, the child was safely de

livered, but at great expense to the perineum. In the other case, a simple forceps delivery was accomplished. but after prolonged labor. In the first of these cases, the patient was of American birth; in the second, of Irish birth; in the third, Irish. Not coming in this series, but almost near enough in point of time to be numbered among these cases, was one of a negress, an American born, having a deviation on the small side of two centimeters in the transverse external measurements, but no deviation from the accepted measurements in the external conjugate. Yet she required a high forceps operation in the delivery of her second child. Her first child was dead born, and, judging from the history given by the mother, craniotomy had been performed. Of the twenty-seven other cases, in this series, without pelvic contraction within my meaning, there were seven having a lesser degree of contraction, but yet not quite measuring up to the regularly applied standard of external measurements. Of these five were in women of Ameri can birth, between twenty-five and thirty-five years of age. One was a primipara of twenty-one, American born, who had a variation on the small side in all the external measurements; the second stage of labor was prolonged with her, but no interference was required. One was in a woman of Irish birth who had borne four children; with all was difficult labor, and in the fifth labor had an occiput posterior presentation requiring forceps and ante-rotation by them. In the labor just previous there had been prolapse of the cord. On the other hand, in one case of the series, that of a primiparous woman, aged thirty-five or more, the external measurements were greater than usual, and her entire labor period, much to my surprise, comprised only seven hours. In all of these cases, owing to the pelvic measurement, I was able to make a fairly good prognosis, as to difficulty or ease of labor. With the limited number of cases here

presented no conclusions can be reached as to the percentage of pelvic contractions in this section of the State, but the findings seem to point toward a state of affairs much as has been found elsewhere. One fact struck me, namely, that among the women between twenty-one and thirty-five of American birth, in nearly every case the measurements were somewhat under the normal standard. This paper is not intended to establish facts, or as a statistical output, but more to emphasize the value of pelvimetry for the general practitioner, if only for prognostic purposes.