now hardly needs any elaborate discussion. When once the mind grasps the idea that a bruit de souffle is due to the regular vibration of a jet of blood passing into a wide space from a narrow one, it appears very unlikely that a similar bruit could be generated by a mere accidental roughening of the interior of the blood channel.

We may now consider, secondly, what evidence there is that the quality of the circulating fluid is concerned in the production of bruit de souffle. Some years ago, De la Harpe,1 of Lausanne, made some experiments on corpses, in which he injected different liquids into the femoral arteries. When he employed oil for this purpose, no bruit was produced; whereas with either alcohol or water a marked murmur was generated. But, as is remarked by Chauveau, in these experiments all the conditions for the production of a sonorous veine fluide were afforded by the passage of the fluid from the canula into the artery. Chauveau himself made a series of experiments on horses, from some of which he withdrew ten kilogrammes of blood by bleeding, while he injected into the veins of other animals ten kilogrammes of tepid water. The state of the blood must have been greatly altered in both cases, yet no bruit was produced.

It is well known, however, that in the human subject an anæmic state is frequently attended by marked murmurs of two distinct kinds-one audible at the base of the heart and along the course of the great arteries, especially perhaps the pulmonary artery; the other (commonly entitled the bruit de diable) heard over the large veins at the root of the neck. Now, it is one of the main recommendations of Chauveau's theory that it professes to explain the occurrence of both these murmurs, the cause of which has hitherto appeared difficult of interpretation.

Chauveau's account of the anæmic murmur heard at the base of the heart and along the large arteries may be stated as follows:He shows, in the first place, that in anæmia the mass of blood is much less than in health. Now, the capillaries, the small and medium-sized arteries, the veins, and the heart itself, with its orifices, are capable of accommodating themselves to this diminution in the quantity of the circulating fluid; but the two main arteries cannot retract in the same degree. Their calibre, therefore, remains proportionately larger than that of the orifices, through which the blood enters them; and the conditions required for the generation of a bruit de souffle are fulfilled. This result is further favoured by the greatly increased force with which the blood is driven into the arteries by the heart in anæmic as compared with healthy subjects, in con

16 Archives Générales,' 1838.

sequence of the great diminution of the arterial pressure, which belongs to that morbid state.

This explanation is certainly ingenious, and perhaps satisfactory; but it appears to us less conclusive than that which Chauveau gives of the venous anæmic murmur, or bruit de diable. Hamernyk long ago pointed out that this remarkable murmur is never heard in any other veins but the cervical; and he seems to have shown that this is due to the fact (first demonstrated by Bérard) that the roots of the innominate veins have peculiar connections, by which they are enabled to resist suction, without being flattened. Their coats are in fact adherent to the cervical fascia over them, and so are fixed to the solid parts around.1 The consequence is that they form on each side a sort of ampulla, the size of which is uniform whatever the state of other vessels. Hence, when the jugular veins retract to meet the diminished mass of blood in anæmia, these ampullæ remain dilated. This of itself satisfies the conditions required for the generation of a veine fluide. But in many cases the narrowing of the veins produced by the anæmia is not of itself sufficient to lead to this result; the application of the stethoscope is then needed to reduce their channel still further. And in some subjects a bruit de diable can be created by the pressure of the stethoscope alone, without their being anæmic or out of health in any way.

Thirdly, it is shown by Bergeon that there is no reason to suppose that a bruit de souffle can be generated, either by increased rapidity of the stream of blood, or by uniform augmentation or diminution of its tension.

It would appear, therefore, that there remains no cause to which the production of a bruit de souffle in vessels at a distance from the heart can be attributed, except the generation of a "fluid vein" in a stream of blood passing from a narrow into a dilated part of the circulating system, in accordance with Chauveau's theory. The conditions under which the blood flows in these vessels are in no way different from those which existed in Chauveau's and Bergeon's experiments, and we are now prepared to admit the universal applicability of their views to all vascular murmurs.

With regard to murmurs generated within the heart and at its orifices, however, we are by no means sure that this explanation holds good to the same extent. There are, indeed, some cardiac murmurs to which it is strikingly applicable. This is the case with the bruits that belong to mitral and aortic stenosis.

1 This "canalization" of these veins (to anglicise the name commonly given to it in French works) renders them especially liable to the dangerous entrance of air during surgical operations, &c.

No one acquainted Chauveau's theory could hear a presystolic bruit, without being struck with the reasonableness of a view that would attribute it to a vibrating jet of blood forced through a narrow mitral ring by an hypertrophied auricle, and impinging on the interior of the ventricle at its apex, close to the very spot where it is heard, and where its thrill can often be so plainly felt.

But for regurgitant bruits (whether mitral or aortic) this theory is less completely applicable. Chauveau himself pointed out, and Bergeon has especially insisted on the same fact, that the murmur caused by a veine fluide is audible only in the dilated space beyond the constriction, and not on its proximal side. Thus, in an experiment, if water be passed through a tube which is narrowed in its calibre for a considerable part of its extent, no murmur is heard, either when the fluid enters the narrowed portion, or when it is flowing through it, but only when it passes into the wider space beyond. Now, it is evident that this applies very well to that part of a regurgitant bruit, which is heard in the direction of the stream; to that part, namely, of a mitral systolic bruit which is heard in the back, near the left scapula; and to that part of an aortic diastolic murmur which is heard below the valves, over the left ventricle and at the apex of the heart. But a mitral systolic murmur is much louder at the heart's apex than in the back; and an aortic diastolic murmur is almost always audible to some extent above the valves. These facts are inexplicable by Chauveau's theory, but Bergeon has supplied a very simple and complete explanation. He has shown that if a tube through which water is flowing be suddenly narrowed in a certain part of its course, and if the narrowed part be made to project backwards into the wider part, so as to form a lip or rim with a circular cul de sac facing the current, a murmur is always produced, which is audible on the proximal side of the obstruction, and transmitted backwards, or against the stream. Now, it is evident that in nearly all cases of valvular incompetency the conditions requisite for the production of such a murmur are fulfilled. The valves, although unable to close, are scarcely ever completely destroyed, but almost invariably project inwards to some extent, and thus form culs de sac, in which the murmur in question can be generated.

This extension, by Bergeon, of the very narrow limits within which Chauveau's theory would have confined the conditions capable of generating the bruit de souffle appears to meet all the facts that require explanation as regards murmurs that are known to be due to obstruction of, or regurgitation through, the cardiac valves. Are we entitled to say that all cardiac

bruits de souffle must be placed in this category? It appears to us that an affirmative answer to this question cannot yet be given. It must be borne in mind that all the physical experiments made by the French writers whose views are under consideration have dealt simply with the flow of liquid through tubes that are either rigid, or at most elastic. No similar experiments have as yet been made, or perhaps could be made, with cavities possessing contractile walls. Is it not very probable that were such experiments to be carried out, it might be found that something more had still to be learnt as regards the physical conditions capable of generating a bruit de souffle?

In certain cases, as is well known, cardiac and vascular murmurs present remarkable modifications in their quality, and we may now inquire to what causes these are due. One of these modifications consists in a peculiar roughness or harshness. Instead of being a soft blowing sound, the murmur resembles the noise produced by a file or rasp, or even by a currycomb. Hence the older French physicians, who devoted much pains to the discovery of the varied morbid changes leading to the development of the murmurs in question, designated them by special names, such as "bruit de scie," " bruit de rape," "bruit d'étrille." If the hand be placed over the spot where one of these bruits is audible, a thrill can generally be felt, such as Laennec termed a "frémissement cataire." The two are so constantly met with together that there can be no doubt of their arising from the same cause; and the question as to the mode of origin of the murmurs under consideration may be taken as identical with that as to the causation of frémissement or thrill.

Now, the latter phenomenon is evidently due to vibrations of the solid tissues, transmitted directly to the surface of the body. It is, therefore, very natural to suppose that these vibrations are its determining cause. According to this idea, an ordinary soft bruit de souffle would differ from a harsh murmur attended with thrill in this-that while the former is due simply to the formation of a veine fluide (as described above), the latter is dependent on the simultaneous vibrations of the orifice at which the veine fluide is generated.

It seems, however, that this opinion is not tenable, although passages might be quoted from the writings both of Chauveau and Bergeon which would, at first sight, appear to support it. Among the points to which Savart directed his attention in the experimental investigations to which we have several times referred, was the question whether the sound produced by a veine fluide is due solely to the vibrations of the liquid or to those of the orifice through which it passes, and he came to the conclusion that the former was the case. For, were it otherwise

"the nature of the material forming the orifice, the existence of an extremely thin edge to it, the degree of polish in the surrounding metal surface, and, above all, the character of the liquid and its temperature, ought to be capable of changing the number of vibrations. But no one of these conditions appears to exercise any influence on the state of movement of the veine fluide, nor is any effect produced when, the orifice being pierced in a brass plate, this is rubbed over with a fatty substance, so as to change completely the state of the surface.

Lastly, if the circumference of the aperture be touched with a solid and resisting body, which would arrest its vibrations (if such existed), or at least affect their amplitude, no change whatever is observed in the state of the vein."

The point, besides its theoretical interest, has some real importance. For, if a rough bruit, with its attendant thrill, be due to the independent vibrations of the solid walls of the orifice at which it is generated, the occurrence of such a bruit may enable us to draw an inference as to the character of the orifice. Now Bergeon, in speaking of the causes of frémissement, says that this can be felt "when the souffle is intense and the tube" (he is referring to simple physical experiments) " thin and elastic. Walls which are thick and rigid do not allow of transmission, whereas supple, thin walls present the most favorable conditions." The use of the word "transmission" in the sentence which we have quoted seems to show that Bergeon is referring rather to the conduction of thrill and murmur from the fluid within than to the development of sonorous vibrations in the tube itself, otherwise we should have understood the passage differently.

It appears to us, however, that the facts of clinical experience fail altogether to support Bergeon's statement. No bruit is so frequently rough and attended with thrill as the presystolic bruit of mitral stenosis. But in this affection the mitral valve, at the orifice of which the bruit is generated, is generally extremely hard, thick, and inelastic; presenting, in other words, conditions the very opposite of those which Bergeon would lead us to expect. But, if we think for a moment, we shall surely have very little difficulty in understanding why the presystolic murmur is so harsh and so commonly accompanied by frémissement. Not only does the veine fluide that generates it flow into a comparatively wide and empty space, but it impinges on the wall of this space, close to the very point to which the finger is applied on the exterior of the thoracic parietes. The full shock of the jet of blood is transmitted directly to the hand of the physician who is examining his patient. It is probable, we think, that in other cases in which frémissement is felt, conditions more