During convalescence the urine increases in quantity, while, as a general rule, the urea and uric acid lessen relatively and absolutely, and the chlorides resume their normal proportions to the other ingredients. The reaction frequently becomes alkaline, and the specific gravity falls considerably, although not always as soon as the articular inflammation subsides. Temporary albuminuria occurs very frequently in the febrile and occasionally in the declining stage, but generally disappears when convalescence is completed. It obtained on admission in 8 out of 43 cases lately reported by Dr. Greenhow.[40] A more abiding albuminuria, due very rarely to acute parenchymatous nephritis, may be met with (Johnson, Bartels, Hartmann, Corm). Blood, even in considerable amounts, has also rarely appeared in the urine,[41] sometimes in connection with embolic nephritis and endocarditis, for such appear to have been the nature of Rayer's nephrite rheumatismale.[42] [Footnote 40: _Lancet_, 1882, i. 913.] [Footnote 41: _Clinical Lectures_, R. B. Todd, edited by Beale, 1861, p. 346.] [Footnote 42: _Traité des Maladies Reins_. See also Dr. Weber, _Path. Trans. of London_, xvi. p. 166.] The saliva, which is normally alkaline, has usually a decidedly acid reaction in acute articular rheumatism, and Dr. Bedford Fenwick states that it always in this disease contains a great excess of the sulpho-cyanides, and that these slowly and steadily diminish, till at the end of the third week or so they become normal in amount.