are required to decolorize the solution, and the proportions necessary will be found to vary from 1 part of urine in 21/2 to 1 in 30 or 40. The subsequent calculation is very simple. If you wish to give the percentage of sugar, multiply 0.005 by 100, and divide the product by the number of cubic centimeters of dilute urine employed. The figure thus obtained, multiplied by the extent of dilution--i.e., if there is 1 of urine in 10, multiply by 10--gives the required percentage. The number of grains per fluid ounce can of course be obtained by multiplying the percentage by 4.375. To observe easily the exact end-reaction a piece of white paper should be placed behind the flask. If the analyst objects to the escape of the waste ammoniacal fumes, they may be conducted by a suitable arrangement into water or dilute acid. In addition to glucose there are small quantities of other copper-reducing bodies present in all urine, which always render the reading higher than strict accuracy would demand. Their aggregate proportion, however, is, comparatively speaking, so minute that for most medical purposes their presence may be disregarded. Greater care must be exercised, though, in those instances where such a deoxidizer as chloral hydrate is accidentally present. In case of doubt, a little washed and pressed yeast should be allowed to stand with the urine for a day or two in a warm place. Alcoholic fermentation with evolution of carbonic acid gas soon sets in, and the specific gravity of the liquid is lowered considerably. This reaction points conclusively to the presence of sugar. Based upon Braun's potassic picrate test, Dr. G. Johnson has devised a colorimetric process for the estimation of sugar. On boiling an alkaline solution of that salt with glucose, the former is reduced to deep red-brown picramate, the color of the liquid, of course, varying in intensity according to the proportion of sugar present. This solution is diluted till it corresponds in tint with a ferric acetate standard, and the percentage of sugar is then readily calculated. For those who prefer this process the convenient apparatus manufactured by Mr. Cetti, of 36 Brooke street, Holborn, is recommended, who will also furnish full particulars of the test. ALBUMEN. Normal urine is free from coagulable proteids, though it is admitted that albumen may sometimes occur in the absence of disease. It is always highly important, therefore, to determine accurately the presence or absence of this body. In the relentless malady named after Richard Bright, the urine always contains albumen, and if accompanied by the "casts" of the uriniferous tubules your report may amount to a sentence of certain death. The tests which we now describe are accurate and easily applied; but reliance should never be placed on any single reaction--at any rate until the operator has acquired considerable experience.