ative result is at once final. A good quality gives a plain reaction with 0.006 per cent. hydrochloric acid, 0.01 per cent. lactic acid, and 0.02 per cent. butyric acid. This blue dye, united with an alkali, is soluble in water, and is sensitive to the hydrochloric acid combined with albumin, this mineral acid having a stronger affinity for the alkaline base of the dye than for albumin. The blue litmus is consequently reddened by all the factors of general acidity.

The second acid qualitative test is made with Congo-red. This is a red dye, soluble in water, and is colored blue by all free acids. It is three or four times more sensitive to free hydrochloric acid than to the free organic acids of the gastric contents. The intensity of the blue coloration increases with the percentage of free acid, and organic acids produce a muddy grayish-blue or purple, in contradistinction to the pure blue of the mineral acid. But even a very extended experience should not permit these slight distinctions to be valued as more than suggestions. The hydrochloric acid and other acids combined with albumin or organic bases, do not alter the color of Congo-red. Acid phosphates, in very concentrated solution, produce a brownish coloration ; but in such quantities as are found in the stomach after a test-meal, produce no change of color. Consequently, the Congo-red is turned blue by the free hydrochloric and the free organic acids. The acid salts and the acid organic combinations are without influence. A positive reaction means the presence of a free acid. Congo-red may be used in aqueous solution or in the more convenient form of paper. The latter reacts plainly with 0.01 per cent. free hydrochloric acid and with three times the quantity of lactic acid. The solution is about ten times more sensitive than the paper, which is made by saturating fine filter paper with an aqueous solution of Congo-red. After it is dry, the paper is cut into narrow strips of convenient length. The test may be made with the solution by spreading a few drops over a white ground, as in a watch-glass, on a piece of white paper, or in a porcelain crucible,-and allowing a drop of the gastric contents to flow over it from the edge; the area of contact becomes blue. Or the paper may be dipped into the mixed contents, after the ordinary manner of using test-papers.

The third qualitative test, in case the Congo-red gives a positive result, is for free hydrochloric acid. The surest and most sensitive reagents for this purpose are those of Günzburg and Boas.

The reagent of Günzburg is an alcoholic solution of phloroglucin and vanillin:

Phloroglucin,
Vanillin,

Alcohol (absolute),

2 gm.

I gm. 30 gm.

This reagent is uninfluenced by all acid salts and combinations and by free organic acids. As hydrochloric is the only free mineral acid present in the gastric contents, unless other mineral acids have been swallowed, a positive reaction with phloroglucin-vanillin is proof of the presence of free hydrochloric acid. This regeant is also very sensitive. With a solution of 1: 10,000 the very fine crystals appear, while a solution of 120,000 gives a red coloration. If there be much organic matter in the contents, the red is mixed with the pasty, dry, yellowish residue. The test is best made by placing three or four drops of the reagent in a porcelain crucible, and spreading it by causing it to flow in different directions over the surface. A like quantity of the filtered gastric contents to be tested is now added, and spread over the same area. With a small flame the crucible is slowly warmed, never allowing it to become too hot to be comfortably borne on the back of the hand. After several seconds a clear red coloration appears, or the fine, bright-red crystals may be seen if free hydrochloric acid be present. The phloroglucinvanillin paper, recommended by some writers, we do not employ. A few drops of the gastric contents are placed on the paper, which may be more rapidly heated in the crucible than with the solution test.

The following solution of Boas' is used in the same manner as the Günzburg reagent:

Resorcin (resublimed),

White sugar,

Alcohol (95 per cent.),

5 gm.

3 gm.

100 gm.

The coloration is a bright rose-red. This reaction of the gastric contents is characteristic of free hydrochloric acid. It is very sensitive, and may be even clearer than the Günzburg reaction when the contents contain much soluble albumin. Dimethylamidoazobenzol, methyl-violet, and tropaolin oo are also employed as free hydrochloric acid reagents.

A very sensitive reagent for HCl is a 1⁄2 per cent. alcoholic solution of dimethylamidoazobenzol. This yellowish solution changes to a reddish color on the addition of a mere trace of free HCl. The reaction is positive with an HCI

solution of 1:20,000. The reaction is also produced by concentrated solutions of acid phosphates and lactic acid, but such concentrations occur so rarely and in such particular conditions as to allow little chance for an error to be made by any one who is familiar with the analysis of the gastric

contents.

Methyl-violet is changed to a sky-blue color by free HCl. The intensity of the blue color produced varies with the strength of the hydrochloric acid solution-1: 10,000 gives a bluish tinge, and 1:5000 a clear reaction. But the reagent is not so sensitive when the gastric contents are employed instead of an aqueous solution of HCl. Lactic acid, in a solution of 1: 300, produces also a bluish tinge. Chlorids also vitiate the results, and the methyl-violet, consequently, is not altogether satisfactory as a free HCl test. The test, which is a very pretty one, is performed in the following manner: A test-tube full of a very dilute solution (1:500) in distilled water is prepared (a clear violet color), and two small test-tubes are about half filled with it. To one is added one to three c.c. of the filtered gastric contents, and to the other the same quantity of distilled water. By comparing the two tubes, the change of color is beautifully displayed.

Tropæolin 00, in concentrated aqueous solution, is an excellent reagent for detecting free acids, but it is not quite so sensitive as Congo-red. The yellow solution is changed to a deep red by free acids. Used in the following manner, it is a sure reagent for free HCl (Boas): Three or four drops of a saturated alcoholic solution are spread thinly in a porcelain crucible, and over the same area an equal quantity of the filtered gastric contents is allowed to flow. Next, heat slowly over a small flame. If free HC is present, lilac streaks appear near the border, which, on further heating, become

blue.

Whenever free HCl is present in the gastric contents, it is useless to make a test for HCl in combination with proteids, since the HCl remains free only when the acid affinities of the proteids have already been satisfied. The presence of HC free proves that HCl in combination with proteids is also present.

But it often happens in the diseases of the stomach that no free HCl can be detected. Under such circumstances, if the reaction of the contents is acid, HCl in proteid combination may be present. To detect it, qualitatively, two tests may be made, one of which is a chemical and the other a color test. A small quantity of the filtered contents is ex

actly neutralized, boiled, treated with acetic acid and sodium chlorid, again boiled, and filtered after cooling. Any albumin left in the filtrate is digested albumin combined with HCl. A positive biuret reaction (rose) on treating the filtered contents with liquor potassa and cupric sulphate shows the presence of propeptones. These chemical tests also give roughly the degree of peptonization. The color test is made by employing a one per cent. aqueous solution of alizarin. Three or four drops of the solution of the dye are added to a small quantity of the filtered contents, and decinormal alkaline solution is added until a pure violet color appears. To a second portion of the filtered contents the same quantity of the alkaline solution is added. If the mixture still reacts acid to litmus, HCl in proteid combination is present. For practical qualitative purposes the biuret test is sufficient.

The Quantitative Analysis.-On account of its bearing on the diagnosis and treatment of the diseases of the stomach, an easy and an accurate method of differentiating and estimating the different factors of the acidity of the gastric contents is very desirable. For a long period the chemical analysis was confined to the detection of free hydrochloric acid. Hayem and Winter (1888) proved the insufficiency of this method, and gave their very valuable, but long, chlorometric analysis. A new light was turned on the chemical pathology of the stomach. Many other quantitative methods have since been given, but none of these seem completely to satisfy the requirements of practice. The busy physician in his daily work demands a quantitative method at once easy, accurate, and rapid.

It would seem that the practical value of very slight quantitative variations in the factors of general acidity has been overestimated, and that the simpler color methods reveal with sufficient exactness all the deviations from the normal chemism which possess a distinctive practical meaning. The inaccurate laboratory methods need not be mentioned. The more complete and chemically accurate methods-which may sometimes be used with advantage in practice, and which should always be employed in original research-will be given. But for daily needs coloration-titration procedures usually suffice.

(a) The Color Methods.-The Method of Mints.-This method estimates the quantity of free HCl, or H, by means of Günzburg's reagent. To ten c.c. of the filtered contents, the decinormal alkaline solution is added from the buret until the reaction of a droplet (platinum loop) of the fluid with the

reagent of Günzburg becomes negative. If the reaction is still positive with 0.9 c.c., but is negative with 1.0 c.c. of the titration alkali, the free HCl acidity in 100 c.c. of the gastric contents is represented by 10 c.c. of the decinormal solution of caustic soda or potash. One c.c. of a decinormal alkaline solution represents 0.00365 HCI; consequently, the 100 c.c. of the gastric contents contain ten times that amount. The strong free HCl in the contents is first completely neutralized by the alkali before any of the other factors of the total acidity are affected. The method consumes less time if the platinum loop, wet in the solution, is brought in contact with Congopaper. When the Congo-red is no longer made markedly blue, the use of the more trustworthy reagent of Günzburg may be begun.

The Method of Boas.-The object of this method is the quantitative estimation of the free hydrochloric acid, or H. Five c.c. of a watery solution of Congo-red are added to an equal quantity of the filtered contents. The mixture becomes blue. The titration is made with the decinormal solution of caustic potash or soda, and contin ueduntil the original Congored color is restored. As a control color, five c.c. of the Congored solution may be added to an equal quantity of distilled water. The titration should be slow near the end, as the restoration of the red color does not take place rapidly.

The value thus found represents free HC1; or, more accurately, free HCl and the quantity of free organic acids present, or HO. The author of the method claims that the quantity of organic acids in the contents after the test-breakfast is practically seldom worth considering when free HCl is present, but that if the organic acids are present in notable quanity, they should be removed by repeated shaking with ether before the titration.

Method of Töpfer.-The method of Töpfer requires the following color reagents: (1) One-half per cent. alcoholic solution of dimethylamidoazobenzol; (2) one per cent. aqueous solution of alizarin; (3) one per cent. alcoholic solution of phenolphthalein.

I. To ten c.c. of the filtered contents are added a few drops of the alcoholic solution of phenolphthalein, and the total acidity is titrated with a decinormal solution of caustic potash or caustic soda. The alkali is to be added until the rose-red coloration is permanent, and drop by drop so long thereafter as the rose color does not become deeper. The end reaction of the color indicator is used. This gives the total acidity, or A.