Admitting with Fuller the occasional deposition of fibrin upon the valves and endocardium in rheumatic fever independently of endocarditis, the murmur resulting therefrom could not be reliably distinguished from that of inflammatory origin. It remains to speak briefly of temporary incompetence of the mitral and tricuspid valves and their dynamic murmurs, and of hæmic murmurs. Occasionally, in severe cases of rheumatic fever, more especially in the advanced stage, there may be heard a systolic murmur of maximum intensity either in the mitral area or over the body of the left ventricle, unaccompanied by accentuation of the second sound, or, as a general rule, by evidence of pulmonary obstruction. Such murmurs are apt to be intermittent, and as they disappear on the return of health, they have been satisfactorily referred to temporary weakness of the walls of the heart, so that the auriculo-ventricular orifices are not sufficiently contracted during the ventricular systole for their valves to close them, and regurgitation follows. Yet, inasmuch as Stokes distinctly mentions the absence of murmur in many cases of softening of the heart in typhus, it is probable that an excessive weakness of the ventricular wall is incompatible with the production of murmur, and that the presence of murmur in such circumstances is evidence of some remaining power in the heart. Dr. D. West[68] has published some cases of acute dilatation of the heart in rheumatic fever which strongly corroborate these views. The murmur in one of them became appreciable only as the heart's sounds increased in loudness and the dilatation lessened. One ended fatally, and acute fatty degeneration of the heart's fibres was found in patches.[69] I believe that some of these temporary mitral murmurs in acute rheumatism depend upon a moderate degree of valvulitis quite capable of complete resolution. Sibson[70] has lately stated that he has met with the murmur of tricuspid regurgitation without a mitral murmur in 13 out of 107 cases of rheumatic endocarditis, and with a recent mitral murmur in 27 out of 50 {36} cases. "The tricuspid murmur generally comes into play about the tenth or twelfth day of the primary attack, along with symptoms of great general illness;" it appears earlier, as a rule, in those cases in which it is associated with mitral regurgitation than when it exists alone; it is of variable duration, but usually short--from one to nineteen days or more. He regards it as of non-inflammatory origin, and dependent upon regurgitation due to the so-called safety-valve function of the tricuspid valve; and when limited to the region of the right ventricle he infers that it is usually the effect and the evidence of endocarditis affecting the left side of the heart. These novel statements are confirmed by the observations of Parrot, Balfour, and William Russell,[71] which go to prove that tricuspid regurgitation occurs frequently in the more advanced stages of debility. No other