[Footnote 50: See Senator in _Ziemssen's Cyclopæd. Pract. of Med._, xvi. 49.] [Footnote 51: _Pract. Med._, 5th ed., 314.] The frequency of cardiac complications in rheumatism is influenced by several circumstances. Some unexplained influence, such as is implied in the terms epidemic and endemic constitution, appears to obtain. Peacock found the proportion of cardiac complications in rheumatism to range from 16 to 40 per cent. during the five years from 1872 to 1876, and a similar variability is shown in Southey's statistical table[52] covering the eleven years from 1867 to 1877. Be it observed that these variations occurred in the same hospitals and under, it may be presumed, very similar conditions of hygiene and therapeusis. Youth predisposes to rheumatic inflammation of the heart, so that it may still be said that the younger the patient the greater the proclivity. Of Fuller's cases, 58 per cent. were under twenty-one, and the liability diminished very markedly after thirty. Of Sibson's cases, 62 per cent. were under twenty-one. In infancy and early childhood the liability is very great, and at those periods of life the heart, and more especially the endocardium, rarely escapes; and the cardiac inflammation often precedes by one or two days the articular. The careful observations of Sibson confirm the spirit, but not the letter, of Bouilland's original statement, and proves that the danger of heart disease is greater in severe than in mild cases of acute rheumatism, and that this is especially true of pericarditis. (It may be remarked here, en parenthese, that the number of joints affected is {33} very generally in proportion to the severity of the attacks.) However, the mildest case of subacute rheumatism is not immune from cardiac inflammation, and it has occasionally been observed even in primary chronic rheumatism.[53] Occupations involving hard bodily labor or fatigue, whether in indoor or outdoor service, render the heart very obnoxious to rheumatic inflammation. Existing valvular disease, the result of a previous attack of rheumatism, favors the occurrence of endocarditis in that disease. Some authorities maintain that treatment modifies the liability to rheumatic affection of the heart, and this will be spoken of hereafter. The period of the rheumatic fever at which cardiac inflammation sets in varies very much, but it may be confidently stated that it occurs most frequently in the first and second weeks, not infrequently in the third week, seldom in the fourth, and very exceptionally after that, although it has happened in the seventh. An analysis of Fuller's experience[54] in 22 cases of rheumatic fever and 56 of endocarditis--a total of 78--shows that the disease declared itself under the sixth day in 8; from the sixth to the tenth in 29; from the tenth to the fifteenth in 17; from the fifteenth to the twenty-fifth in 18; and after the twenty-fifth in 6. The friction sound was audible in Sibson's 63 cases of rheumatic