point, the projection lifts the fork until the platinum points dip into the mercury, when a new contact is made and the current is sent through the opposite electromagnet. This now pulls the frame and shuts off the gas-flow. As soon as this contact is broken, the minimum oxygen delivery (which has been originally adjusted) is once more established. It will be observed that very little electric power is wasted in this process, as the current passes only at the time of the contact, which, in either case, is immediately broken.

In Plate 2 the entire apparatus is shown with the electrical connections. I use the street current directly and, by the use of a telephone condenser of 2 micro-farads (1) across the binding posts of each magnet and one ordinary lamp (2) in series, the contacts are made without any sparking. As the circuits are used alternately, the two magnets are supplied from the same socket (7).

While the apparatus here described has a definite purpose, it is evident that it may be employed with advantage for other functions. It would be an easy matter, for instance, to utilize the expansion of mercury for making the contacts and thus regulating the flow of lighting gas for the heating of incubators or drying ovens.

I.

SANITARY STUDIES OF BAKING POWDERS

Is aluminium absorbable from bread, and similar food products, made with alum baking powder?

WILLIAM J. GIES

(Biochemical Laboratory of Columbia University, at the College of Physicians

and Surgeons, New York)

(Received for publication, April 1, 1916)

Five years ago I published a paper, entitled "Some objections to the use of alum baking powder,"1 from which the introductory paragraph is quoted below:

"During a period of about seven years I have occasionally conducted experiments on the effects of aluminium salts. These studies have convinced me that the use, in food, of alum or any other aluminium compound is a dangerous practice. That the aluminium ion is very toxic is well known. That 'aluminized' food yields soluble aluminium compounds to gastric juice (and stomach contents) has been demonstrated. That such soluble aluminium is in part absorbed and carried to all parts of the body by the blood can no longer be doubted. That the organism can 'tolerate' such treatment without suffering harmful consequences has not been shown. It is believed that the facts in this paper will give emphasis to my conviction that aluminium should be excluded from food."

Among the experiments above referred to, as having been conducted "during a period of about seven years," were those with House, on seedlings, and with Steel, and with Kahn, on dogs.

The gist of our conclusions, from the results of these experiments, is indicated by the following four quotations:

I. "Compounds of aluminium, such as the chlorid, even when present in very minute proportions, are strong protoplasmic poisons and very toxic to growing plants" (House).

II. "When alum was administered in aluminium-free food to dogs, or when dogs ingested biscuits baked with alum baking powder, aluminium in comparatively large amounts promptly passed into the blood. When aluminium chlorid was administered intravenously, from 5.55 percent to 11.11 percent of the aluminium passed from the blood into the feces during the three days immediately after the injection" (Steel).

...

III. "When biscuits baked with alum baking powder are fed in a mixed diet to dogs, aluminium passes in considerable amounts into the blood. Such absorbed aluminium circulates freely and, although it does not show a tendency to increase proportionately in the blood, it accumulates to some extent in various parts of the body. The bile contains a particularly large amount of aluminium. Under the conditions of these (feeding) experiments, aluminium is absorbed in part and is excreted, to some extent, in both the bile and urine" (Kahn).

IV. "Two objections might be raised against this conclusion regarding the absorption of aluminium from the gastro-intestinal tract into the blood. It might be assumed that our analytic method was inaccurate. Such an assumption is always in order in the consideration of any method, for none is perfect. Our preliminary work with the analytic method employed and the checks against error that were instituted, give us complete confidence, however, in the practical accuracy of the process. If slight and unavoidable errors occurred, we believe that, from the general nature of the method, they must have caused loss of aluminium and consequent diminution of the numerical result, rather than the reverse.

"It might also be urged that our figures representing the amounts of aluminium which passed into the blood are too small to mean anything. It might be thought that the proportions are too minute for the causation of deleterious effects. It might even be suggested that such small proportions of aluminium are beneficial because of the possible 'stimulating' action which traces of poisons frequently induce. But

For further facts in this connection see Gies: BIOCHEMICAL BULLETIN, 1916, v, p. 189.