were transferred to a desiccator before weighing. It was found that in most cases no additional loss of weight resulted on leaving the samples in the oven more than 14 hours.

The desiccators employed in the vacuum desiccator method were of the Scheibler form, 6 inches in diameter and of strong glass with a rubber stopper carrying the stop-cock wired and sealed in place. About 150 cc. of sulphuric acid were placed in the bottom of each desiccator and an inverted glass stopper of the hollow type containing 10 cc. of absolute ether was placed in the acid below the rack which held the aluminium dishes. The samples, also a mercury manometer, were placed on the racks of the desiccators and 15 minutes after connecting the house vacuum system to the desiccators all the air was removed from the desiccators. The vacuum was made complete, as shown by the manometer in each desiccator, by carefully shaking the desiccators to insure the absorption of the remaining traces of the ether fumes by the sulphuric acid. In order to secure rapid absorption of moisture, the desiccators were shaken every few hours to break the film of moisture which settled on the surface of the sulphuric acid. The samples were allowed to remain in the desiccators for 5 days when they were weighed and replaced in the desiccators for an additional period of 48 hours. In practically all cases, the first 5-day period was sufficient to remove all moisture that could be removed by this method.

COMPARISON OF RESULTS.

In Table 1, results for moisture in powdered malted cereal preparations are given by the two methods. The samples which were commercial products consisted largely of sugars or starch and protein with a little fat.

Samples 1, 4, 5, 6, 7 and 8 were composed of a large percentage of sugars and contained but little starch. These samples gave a larger loss of weight (moisture) on heating in the vacuum oven at 65°C. than in the vacuum desiccator. The other samples which contained large amounts of starch and but little sugars gave a greater loss of weight (moisture) in the vacuum desiccator than in the vacuum oven at 65°C.

Samples 1, 2, 3 and 4 were heated an additional 4 hours in the vacuum oven at 100°C. with the following results:

Table 2.

Samples used for results in Table 1 heated 4 hours additional in vacuum oven at 100°C.

These samples were then returned to the vacuum oven for 6 hours' additional heating at 100°C. Samples 2 and 3 showed an increase of weight while Samples 1 and 4, which were low in starch and high in sugars, did not show an increase of weight.

In Table 3, similar results for moisture determined by the two methods on four samples of malted milks are given. These products contained a large amount of sugars, dextrins, etc., and practically no starch. In all four samples a greater loss of weight resulted on heating in the vacuum oven at 65°C. for 14 hours than in the vacuum desiccator method.

The samples which had been in the vacuum desiccator were placed in an oven at 100°C. for 4 hours and the samples which had been heated in the vacuum oven at 65°C. were returned to the same oven for 4 hours' additional heating. The results are as follows:

While the samples heated at 65°C. all gave higher results, there was but little difference except in the case of Sample 4.

The samples, the analyses of which are reported in the foregoing tables, contained such a low percentage of moisture that the influence of the various conditions of the methods employed may be magnified in the results, and, therefore, appear to a greater extent than with samples higher in moisture. Samples of practically identical moisture content and containing both sugars and starch, and starch products in considerable amounts, do not show these differences in moisture results (loss of weight) by the two methods. There is no certainty that the higher figures are the correct moisture results in any of the samples tested.

The results obtained by determining the loss of weight or moisture in pure proteins, fat, starch and various sugars can hardly be applied to these products, since they are complicated mixtures of these substances, and are the result of various methods of manufacture. The carbohydrate composition of the cereal products tested showed a greater influence on the moisture results (loss of weight) than the moisture methods employed.

In Table 5 results are given by the two methods for moisture in veal, fish and meat extracts. With veal and fish, the results were practically

identical by the two methods, while a greater loss of weight resulted with meat extracts by the vacuum desiccator method, than on heating at 65°C. in the vacuum oven.

Meat and fish are complex organic substances containing protein, fat and only a little carbohydrate. The fact that identical moisture results were obtained by the two methods may be explained on the basis of the low carbohydrate content of the veal and fish. The high percentage of moisture, viz, 75 per cent, masked any variations which were caused by the small percentage of carbohydrates present. In the case of meat extracts, Table 5, the physical character of the samples played an important role, as the large expansion of the samples which took place in the vacuum desiccator facilitated the removal of the moisture and, therefore, may explain the higher results obtained in the desiccator method.

SUMMARY.

A difference in the moisture content (loss of weight) of malted cereal preparations high in starch and similar preparations high in sugars by the vacuum desiccator method and by the method of heating at 65°C. for 14 hours in a vacuum oven was obtained.

Those samples high in starch gave lower moisture results and those high in sugars gave higher results on heating in a vacuum oven at 65°C. for 14 hours than by the vacuum desiccator method.

With meat and fish, which are practically carbohydrate free, identical results were obtained by the two methods. Meat extracts gave a higher moisture result (loss of weight) in the vacuum desiccator than in the vacuum oven at 65°C.

COMMERCIAL FEEDING STUFFS.

By A. MCGILL (Department of Health, Ottawa, Canada).

So far as Canada is concerned the Commercial Feeding Stuffs Act of 19091 has for its object such control of all feeds for cattle as shall assure, to the purchaser, a determinate minimum food value in the article which he buys. This, of course, assumes that he buys intelligently, i. e., that he knows what he wants, and demands this under a definite name, understood by all the parties to the transaction. Feeds for farm stock are either subject to legal inspection, or are permitted to be sold without inspection.

Hay, straw, roots, wet brewers' grains, as well as oats, barley, corn, etc., in the unground state, are not subject to any official inspection.

1 Statutes of Canada, 1909, ch. 15.

The reason is clear. These articles are, from their nature, as readily judged for quality by the ordinary purchaser as they could be by a government official.

All other feeds are subject to inspection for the equally intelligible reason that their value is neither readily nor certainly appreciable to the purchaser. Indeed, the only way in which the superiority of one article of bran or chop feed over another article of like species can be found, without chemical analysis, is by actually feeding the material and observing the results. Feeds which are subject to inspection are again divided into two classes: Those which may be sold without registration; and those which must be registered before being legally offered for sale.

Bran, shorts (middlings) and chop feeds are not required to bear any registration number, or any guaranty of their nutritive value. This is because the terms bran, shorts and chop feeds are legally defined. The feeds in question are required to possess the subjoined percentage values: Standards for bran, shorts and chop feeds.

The purchaser of any one of these feeds is assured of obtaining a nutritive value at least equal to the value above given.

There is, however, nothing in the act which prevents a manufacturer of bran, shorts, or chop from registering his product, and from guaranteeing a higher value for it than the minimum value required by the act. Should a manufacturer of bran, etc., elect to register and to guarantee a value higher than the minimum which the act requires for such class of feed, it is incumbent upon such manufacturer to live up to the guaranty set by himself. If, for example, he chooses to warrant his bran as containing 16 per cent of protein, and 5 per cent of fat, but puts on the market an article containing only 14 per cent of protein and 3 per cent of fat, such article will be adulterated, under the act, although it meets the minimum requirements for legal bran.

All feeds other than those described above must be registered before being offered for sale. The registration number must be affixed to every package offered for sale, and constitutes identification of the brand. Every registration number corresponds to a definite brand of feed, whose minimum value in nutrient matter is guaranteed by the manu