An exceedingly delicate and convenient test is that by nitric acid. The acid is placed in the bottom of a conical wine-glass, and the urine, filtered if necessary, allowed to flow on top of it from a pipette, so as to disturb the plane of junction of the two fluids as little as possible, and leave a distinct line of demarcation. At this plane of union, if albumen be present, will be formed an opaque white line varying in thickness according to the amount of albumen, so that after some practice and with care an approximate estimate of the percentage may be made. A deposit of urates may sometimes be formed a little above the plane of union, but it may be distinguished by its position, by its less distinct limitation on the upper surface, and also by its disappearance on warming. In a very concentrated urine and in cold weather this error may be conveniently avoided by previous warming of the urine and of the reagent. The same remark applies to the brine test. A crystalline precipitate of nitrate of urea may give rise to error if the urine be very concentrated or the experiment conducted in the cold. This may be distinguished by its disappearance on warming or by the microscope. The action of the nitric acid on the coloring matter of the urine, forming a dark band at the point of junction, may obscure the reaction, but with care will not give rise to mistakes. Another test recently introduced, which presents some advantages over the nitric acid, and is certainly quite as delicate, consists in a saturated solution of common salt in water acidulated with about 5 per cent. of the dilute hydrochloric acid of the _Pharmacopoeia_. This solution should be used exactly in the manner described for nitric acid. There is no change of color at the line of junction, and no precipitate takes place there except albumen or peptone, or resins when they have been administered. The opaque line of precipitate may, if the amount of albumen present be small, require a short time to form, so that in cases of doubt it is well to allow the test-glass to stand for a few minutes. It will, however, show very distinctly in any cases in which nitric acid shows any precipitate. The line does not, however, increase in thickness and density in proportion to the amount of albumen so exactly as that produced by nitric acid, so that the brine test is not so useful for approximately quantitative use as the nitric acid, although fully as delicate. If it be desired to distinguish peptone from albumen, it may be done by a comparison of this test {37} with the nitric acid, which does not throw down peptone. If a deposit occur, which may consist of resin, the addition of more urine will dissolve it if resin, while albumen will not be affected.