men predominated in the usual proportion. My method was to give the drug in capsule form, and in doses ranging between four and five grains. Fortunately, my patients could swallow capsules, so that the administration was easy. I ordered it taken about two hours before eating, so that the solution would have an opportunity to exert a soothing and healing effect on the sore. The results as I noted them were as follows: the hyperacidity came under control in about forty-eight hours; the severe distress ceased in one day; patients who walked the floor at night were able to sleep undisturbed almost from the beginning of treatment. The localized soreness usually persisted for about ten days. The patients improved in spirits, and quickly regained strength and color. I usually kept up the treatment for two weeks, sometimes three weeks. I was guided entirely by the relief of the symptoms. The more recent the case the quicker did I get results. Where the history indicated a first attack of recent date the relief of symptoms was very prompt. Where, however, the case was one of several recurring attacks going over a period of several years, as in one of my cases, it required from a week to ten days before the patient was relieved of his symptoms.

As to the permanency of the cure, if I may use that word, I can not say as yet. We know that duodenal ulcer is chronic and recurring over a long period of years. The average duration of the Mayos' cases was over 12 years, and freedom from attack from two years after operation is regarded by them as a cure. My cases only go back about one year, but during this time there has been no recurrence so far. It will take more time and more cases to prove this point. So far, all that I can say is that this method of treatment is perfectly safe, there has been no argyria in any of the gastric or duodenal cases that I have treated, and the results, as far as the relief of symptoms and prompt return to

normal health are concerned, far surpass any other method that I know of.

CORN OIL.

W. L. Cummings, in the Physicians' Drug News, tells us that this product is made from the seed of Indian corn (Zea Mays). There are two processes of manufacture: (1) in which the seed is pressed before it is used for the manufacture of starch, which produces an oil of a golden yellow color and (2) where it is recovered from the residue of the fermentation vats where the corn has been used in the production of alcohol.

This oil is almost without drying power, neither boiling nor the addition of lead when boiling giving it definite drying properties. If it is heated, however, and a current of air passed through it, and manganese borate mingled with it, it dries after a fashion. It is largely used at present in the manufacture of what are known as corn-oil rubber substitutes. It also has a limited application in soaps, such as the soft soaps used in washing automobiles, being cheaper than linseed oil for this purpose and giving practically as good results. In my own experience, says the author, I have found it admirably adapted to this purpose, either alone or in combintion with resin. These soaps may be loaded quite heavily with silicate of soda and when properly made are of a transparent amber color, possessing excellent cleansing properties.

Corn oil was proposed at one time as a suitable material for use in making commercial linseed oil substitutes, but its lack of drying power renders it less suitable for this purpose than various other materials. It is, however, used to a limited extent for adulterating linseed oil.

For most of the purposes to which corn oil has been put, soja bean oil provides a still cheaper and quite as satisfactory material. Practically the only exception to this rule is for the production of rubber substitutes.

Caffeine Tannate as the Chief Con

T

stituent of Tea Infusion.

HE view that caffeine exists in an infusion of tea in the form of a definite compound with tannin is sustained by further experiments. In a previous article upon the subject* it was shown that when an infusion of tea (which always exhibits an alkaline reactien) was acidified with any acid a flaky buff-colored precipitate settled out, especially in the cold, which proved to contain chiefly caffeine tannate. Further, the proportion of caffeine to tannin in this precipitate was found to be 1 of the former to 3 of the latter, and the suggestion was that the compound consisted of 1 molecule of caffeine CH,NO, with a molecular weight of 194, associated with 1 molecule of quercitannic acid CHO, with a molecular weight of 602. In the previous article the formula was wrongly given as CHO,. The ratio of caffeine to tannin in such a compound would be 1:31. The following is an example of the analysis of this precipitate from a good Indian tea:

Caffeine

Tannin

Resinous and oily matters (by difference)

Total...

20 20

20.93% 60.80% 16.27% 100.00%

The ammonium sulphate precipitate amounted to 4% of the tea used in the infusion (5 grams in 400 cubic centimeters of just boiling water allowed to stand five minutes before pouring off). Calculation shows that this particular tea yielded a total of caffeine tannate of 12-26%. Of this total 8.26 was thrown out by mere addition of acid and 4.00% by subsequent saturation of the filtered acid infusion with ammonium sulphate. There is no point in separating caffeine tannate first by acidifying and next by saturating with ammonium sulphate. The whole can be thrown down at once by adding ammonium sulphate after acidifying the tea infusion with dilute sulphuric acid. When this precipitate is air-dried and sunk in a mixture of 1 of alcohol and 2 of benzene the caffeine tannate is separated from the excess of ammonium sulphate crystals and resins and obtained in a comparatively pure state. The solution of caffeine tannate in benzene-alcohol mixture readily yields its tannin to lead oxide (litharge), the caffeine being set free. To give an example, a certain tea (Indian) yielded a total precipitate by ammonium sulphate amounting to 11.60% when dissolved in

benzene-alcohol mixture and evaporated to dryness and weighed. The following was the process: A definite volume of the alcohol-benzene solution of the caffeine tannate was evaporated over a weighed quantity of lead oxide. The dry residue was then made into a thick paste with water and extracted with chloroform, which was poured off and evaporated in a platinum basin. The residue from the chloroform represents the caffeine contained in the caffeine tannate. The wet lead oxide paste was then dried to constant weight and the increase of weight shown over the original' amount of lead oxide weighed out was regarded to be due to tannin. In this way a residue of 11-60% obtained by evaporating the alcohol-benzene mixture to dryness gave on treatment in the way just

Total (caffeine tannate).. 12.80% These experiments thus appear to establish that, complex as the composition of an infusion of tea may be, it at least yields a definite substance in the shape of caffeine tannate, the composition of which is constantly shown to be one part of caffeine with three parts of tannin. The Preparation and Composition of Caffeine Tannate.

Caffeine tannate is a very interesting substance apart from its occurrence in tea infusion. It may readily be prepared by mixing aqueous solutions respectively of tannin and caffeine in the cold, when a milky fluid, sometimes with flakes forming according to the strengths of the solutions, is obtained. If a few drops of

carbonate of soda solution be added to the mixture a clear pale yellow solution results. If next a slight excess of dilute sulphuric acid be added caffeine tannate separates in flakes which admit of washing to remove free caffeine or tannin as the case may be. These flakes may be dissolved in alcohol-benzene mixture

(1 to 2). On evaporating the solution to dryness a crisp crust of caffeine tannate is obtained. If this residue is taken up again in benzene-alcohol and evaporated over a weighed quantity of lead oxide the two constituents are separated. On evaporating the benzene-alcohol and moistening the mass with water to a paste chloroform may be added which will extract the whole of the caffeine; the lead residue on drying to constant weight will give the amount of tannin as an increased weight on the lead oxide employed. To give an illustration: A benzene-alcohol solution of the precipitate obtained by adding acid to a solution of caffeine and tannin in weak carbonate of soda was evaporated to dryness. The weight was 0.1 gram. The residue was redissolved in benzene-alcohol and the solution so obtained was evaporated to dryness in a platinum basin containing 3.210 grams of lead oxide (litharge). The dried mass, after moistening, was extracted with chloroform. The residue (caffeine) from the chloroform, clean and white, weighed 0.025 grams. The lead oxide, after being exhausted with chloroform, was then dried at 120° C. to constant weight. The weight proved to be 3.287 grams, which represents a gain due to tannin of 3.287-3-210, or 0.077 grams. The 0-100 grams of caffeine tannate was thus split into 0-025 grams of caffeine and 0-077 of tannin, figures which account for the whole of the original substance taken, and which give the product 3.08 when the proportion of tannin is divided by the proportion of caffeine. Caffeine tannate is readily soluble in boiling hot water, but as the solution cools it turns milky. It is, however,

permanently soluble in water containing a little alcohol. If caffeine tannate be heated with a little water for some time it assumes the form of an oil.

It was next decided to add an excess of caffeine in solution to a known amount of tannin in solution. If tannin selects a definite amount of caffeine to form a compound it is evident that by estimating the remaining free caffeine in the mixed solution the amount that has combined with the tannin can be ascertained. Thus, 0-3 grams of caffeine in solution was added to 0-3 grams of tannin in solution, a few drops of carbonate of soda solution were added to make the mixture quite clear, and then dilute sulphuric acid was added in slight excess. Caffeine tannate readily separated in flakes from which a clear fluid could easily be poured off. The clear fluid was then repeatedly extracted with chloroform and the chloroform extractions were evaporated. The data and results were as follows:

Calculated excess of tannin= 0.21 gram. The mixed solutions were saturated with ammonium sulphate and thrown upon a filter. The filtrate was exhausted with ethyl acetate and the extraction evaporated to dryness (we noted before that free tannin is soluble in saturated ammonium sulphate, but that caffeine tannate is insoluble). The ethyl acetate gave 0-205 gram free tannin against 0.21 calculated. The caffeine tannate formed was again split up by means of lead oxide, and there were obtained 0.10 gram tannin and 0.03 caffeine, a total of 0.13 gram caffeine tannate formed against a calculated 0.12 gram.

Caffeine Tannate Physiologically Con.

sidered.

We suggest that these findings are important from a physiological point of view, because it is highly probable that caffeine in a state of stable chemical union with tannin must differ very materially in regard to therapeutic action. from the alkaloid in a free or uncom> bined condition, especially having regard to the fact that tannin serves to some extent as an antidote to most alkaloids, since it forms with them insoluble compounds. Similarly, tannin when combined with caffeine is a totally different thing from tannin in the free state. The astringent properties of tannin, for example, disappear when it is combined definitely with caffeine. It follows that tannin in combination with caffeine will not tan. A solution of pure caffeine tannate does not act upon gelatin or proteins, so that when tannin is combined with caffeine its incompatibility with these dietetic substances ceases, and its objectionable retarding action upon digestion accordingly must be modified. If these conclusions are correct it becomes of the utmost importance to determine whether in the enormously popular bev

erage tea, caffeine or tannin exists in the free state or mutually combined.

We have already shown that to a large extent the teas which are regarded as of high quality by the merchant prove to yield infusions containing caffeine and tannin in the ratio of 1 to 3, and the strong inference is that such infusions contain neither free tannin nor caffeine, but neutral caffeine tannate. The merchant's view of quality would thus appear to accord quite happily with physiological considerations. He is guided, of course, chiefly by the flavor of the infusion, as well as by color and odor and general appearances. Both tannin and caffeine in the free state have characteristic tastes; the former is astringent and sour, the latter is bitter. A slight excess of caffeine in tea infusion is probably less objectionable than an excess of tannin, since the disagreeable qualities of tannin are more marked. An ideal infusion is one which contains both caffeine and tannin in 1 to 3 proportion, i. e., caffeine tannate, neither constituent being in excess. This ideal is occasionally reached, but, generally speaking, most good teas contain caffeine in slight ex

cess.

Caffeine tannate, however, does not possess the qualities of its constituents; it has a peculiar flavor which is smooth and bland, not unlike a very delicate or lightly infused tea. It must be carefully borne in mind that in making these deductions we are not considering the chemistry of the leaf, but of its infusion made by pouring 400 c.c. of just boiled water upon 5 grams of tea and decanting after five minutes-a plan suggested by the tea-taster's method of assaying tea.

The Effect of "Stewing" Some Teas.

The desirable equilibrium may be disturbed by vicious treatment even of a good tea. If an Indian tea be boiled long enough or stewed on the hob a point is reached when tannin appears over and above the amount which can combine

with the caffeine to form neutral caffeine tannate. Such an infusion presents the objectionable characters of tannin. When this infusion is saturated with ammonium sulphate the caffeine tannate as such is thrown out. If this is filtered off the excess of tannin will be found in the filtrate, from which it can be extracted by ethyl acetate. Thus the five minutes' infusion of an Indian tea showed a total tannin content of 9-24% and caffeine 3.70%. On infusing the same tea for one hour the tannin amounted to 16-12% and the caffeine 4-40%. In the five minutes' infusion, therefore, caffeine was in slight excess (0-38%) of the tannin, but in the one hour's infusion the tannin is in excess of the caffeine by 2.92%. In other words, 4-40 of caffeine if combined as tannate would mean in the combination 13-2 of tannin, when as a matter of fact the total tannin found was 16-12%.

The experience of making wholesome tea seems to have decided upon a limited time for making the infusion. This limit. in fact, appears to us to insist upon the infusion consisting as far as possible of caffeine tannate and upon the exclusion of an important quantity of free tannin or caffeine present. But even a five minutes' infusion of some teas may not approach this standard. Indeed, in view of our work upon the subject we are able to suggest a classification for teas as follows:

Good teas are those which on a five minutes' infusion yield only caffeine in the form of caffeine tannate to the infusion, neither caffeine nor tannin being in excess; and bad teas are those which yield on infusion for five minutes a tea containing in addition to caffeine tannate either caffeine or tannin, but especially tannin, in excess.

It is obvious from this classification that a good tea may be made a bad tea, but a bad tea can not be made a good tea, except possibly by very skillful blending. Excessive infusion will spoil a good tea, but a short infusion of a bad tea may be