Excretion of chlorids is diminished in nephritis and in fevers, especially in pneumonia and inflammations leading to the formation of large exudates. In nephritis the {57} kidneys are less permeable to the chlorids, and it is probable that the edema is due largely to an effort of the body to dilute the chlorids which have been retained. In fevers the diminution is due largely to decrease of food. In pneumonia chlorids are constantly very low, and in some cases are absent entirely. Following the crisis they are increased. In inflammations leading to formation of large exudates--_e.g._, pleurisy with effusion--chlorids are diminished, because a considerable amount becomes "locked up" in the exudate. During absorption chlorids are liberated and appear in the urine in excessive amounts. [Illustration: FIG. 16.--The Purdy electric centrifuge.] _Quantitative Estimation_.--The best method for clinical purposes is the centrifugal method. {58} _Purdy's Centrifugal Methods_.--As shown by the late Dr. Purdy, the centrifuge offers an important means of making quantitative estimations of a number of substances in the urine. Results are easily and quickly obtained, and are probably accurate enough for all clinical purposes. [Illustration: FIG. 17.--Water-motor centrifuge.] In general, the methods consist in precipitating the substance to be estimated in a graduated centrifuge tube, and applying a definite amount of centrifugal force for a definite length of time, after which the percentage of precipitate is read off upon the side of the tube. Albumin, if present, must be previously removed by boiling and filtering. Results are in terms of _bulk of precipitate_, which must not be confused with _percentage by weight_. The weight percentage can be found by referring to Purdy's tables, given later. In this, as in all quantitative urine work, percentages mean little in themselves; the actual {59} amount eliminated in twenty-four hours should always be calculated. The centrifuge should have an arm with radius of 6¾ inches when in motion, and should be capable of maintaining a speed of 1500 revolutions a minute. The electric centrifuge is to be recommended, although good work can be done with a water-power centrifuge, or, after a little practice, with the hand centrifuge. A speed indicator is desirable with electric and water-motor machines, although one can learn to estimate the speed by the musical note. [Illustration: FIG. 18.--Purdy's tubes for the centrifuge: _a_, Percentage tube; _b_, sediment tube.] * * * * *