If doubts have been expressed as to the fact of iron absorption, there has been still more controversy as to the mode in which it is effected, and this, indeed, may differ according to the preparation employed. The finely divided metal, "reduced iron," is first oxidized by the help of water (for if the compound contain any sulphur, disengaged hydrogen makes itself evident as a sulphuret in eructations). The protoxide and the carbonate, themselves not soluble enough for absorption, are rendered so by the hydrochloric acid of the gastric juice (as evidenced by experiments with the gastric juice of dogs); the protochloride does not coagulate albumen, and is readily absorbed. The sesquioxide becomes first perchloride and then protochloride, and is absorbed as such (Rabu-teau).

It was formerly held that all proto-salts became quickly changed in the system into per-salts, because this change so readily occurs outside the body, but various conditions will prevent or even invert it. Thus, Quevenne notes that a natural protocarbonate remains as such in many mineral waters - that per-salts are reduced by alkaline tartrates, by charcoal, or simply by cold - and that the ethereal "tincture of steel" is rendered colorless and reduced to protochloride by mere exposure to the air and light. Stenhouse found the per-salts to be reduced by organic substances generally, and 0. Bernard, after injecting a per-salt into the jugular vein, recovered only a proto-salt from the urine. In short, it is probable that if iron exists at all in the system as a per-salt it is only for a time, and under the temporary influence of an increased amount of oxygen; its rapid change from one condition of oxidation to another is possibly in accord with a general law of the organism (Quevenne).

Salts of the organic acids, the citrates, lactates, and tartrates, may be absorbed directly into the blood, the acid becoming quickly oxidized or

"burnt off," and the metallic base left free to combine with the blood constituents. Rabuteau suggests that a carbonate of iron may be formed, as are carbonates of the alkalies after administration of alkaline citrates, etc.

The potassio-tartrate or tartarized iron has seemed to be more readily assimilated than any other preparation (Leras, Mialhe). The iodide of iron exhibits the properties of iodine rather than of the metal, and has proved more irritating than simple iron compounds. The whole of the iodine has been found eliminated in the urine after a few days, while but little of the iron has passed out (Quevenne, Melsens), proving that complete separation of the constituents occurs in the system.

Salts of the mineral acids - the chloride, nitrate, and sulphate - if given so diluted as not seriously to constringe the gastric membrane, nor to coagulate albumen, may be absorbed directly into the blood, and much more quickly than the metallic preparations. (Mialhe suggests that the blood-alkalies combine with, or neutralize, the acids so as to leave the metal free for oxidation or combination.) Stronger solutions must, for their first effect, coagulate the albuminous material they meet with in the stomach. Mitscherlich found a "proto-albuminate" of iron in the stomach of rabbits; but, within certain limits, this compound is soluble both in excess of the iron solution, and in fresh quantities of albumen. Gubler, indeed, recommends it as a good form for administration, being both active and non-irritant. It has been generally held that this compound of iron was absorbed and circulated as a per-albuminate, but, according to Dietl, proto-albuminates are much more soluble than the per-salts (Schmidt's Jahrb., 1874); the same has been said of proto-chlorides, but practically both forms are available. The precipitates formed by per-salts with albumen are soluble under various conditions; using 1 1/2 part of ferric chloride to an albuminous solution, H. Rose found the precipitate dissolve in an excess of the salt, and when quite fresh, even the "blood-alkalies" dissolved it. When albumen came into contact with weak iron preparations a few drops of weak acid, or sometimes of alkali, were enough to help solution in the gastric juice (Dietl). As qualifying the observation of Lersch that albumen could supply the place of acids as a solvent for iron in the gastric juice, Dietl found that iron albuminate was soluble in soda solutions, that iron phospho-albuminate was soluble in contact with phosphate of soda, and that alkaline phosphates generally favored the absorption of iron salts after they had become albuminates (Schmidt, loc. cit.).

It is probable that iron is not only absorbed as an albuminate, but eliminated mainly by membranes having albuminous secretions, such as mucous and serous membranes. Dr. Ringer refers to experiments in which, after being injected into the blood, most of the iron was detected on the mucous lining of the intestine, the bronchi, the gall and urinary bladder, and the serous membrane of pericardium, etc. Gubler relies on such facts to explain the astringent action of iron on various parts distant • from the stomach, reasoning that the metal becomes separated by such secreting surfaces from the albumen with which it has been combined, and then recovers and exerts its natural astringency (cf. vol. i., p. 228).

To resume: of the three groups of preparations - the first comprising the reduced metal and carbonate; the second, astringent acid soluble compounds; and the third, soluble non-astringent ones - all are susceptible of absorption under favorable conditions, the first comparatively slowly, the second (when diluted) quickly, and the third group to a medium extent. This fact becomes of importance in guiding our choice of a preparation in different maladies. Of either group a certain proportion, according to the dose, the condition of the stomach, etc., may remain unacted upon, and consequently unabsorbed, and pass into the intestine mechanically mingled with the food; at this stage some further proportion is absorbed under the influence of fats (Mialhe), or of alkaline secretions, or of bile. It has long been recognized that the bile contains a relatively large proportion of iron, and Lusanna argued that the greater part, if not all, of the administered metal passed only into the portal circulation from the mesenteric vein, and was eliminated by the bile. More recently the fact has been used as an argument to show that the effete blood-corpuscles are broken up in the liver, and furnish to the bile its large proportion; thus, Dr. Young, after many analyses, fixing .0065 as the amount of iron contained in the 100 grammes of human bile, calculated it to represent 6.63 grammes of corpuscles (Journal of Anatomy and Physiology, 1871). However it be, iron is largely eliminated in that secretion, but any amount that passes through the intestine is liable to be changed into tannate or sulphide, and so to color blackish the fecal mass; if this be of ordinary consistence, its external surface, which is alkaline, will be found more deeply stained than the inner part, which is slightly acid (Quevenne). In suckling children, the coloration will not occur, and if the salt be very completely absorbed, as are small doses of proto-chloride (Rabuteau), or of tartrate (Stille), it will not be noticed for some days, or until the system is saturated. Again, if the iron pass without any assimilation, it is also said not to color the stools, so that this effect has been considered, though I believe incorrectly, some guide to the absorption of the drug (Kraus, in Ranking, i., 1872, p. 272).