But were it much less difficult in every case accurately to separate the constituents of gluten, it would not, in my opinion, be of the least practical utility. It is to the peculiar mechanical property of this gluten that wheat flour owes its superior power of detaining the carbonic acid engendered by fermentation, and thus communicating to it the vesicular spongy structure so characteristic of good bread.[37] It may also be added, that the results of more than one hundred trials have satisfied me that a diminution or loss of elasticity in the gluten is the surest index of the amount of injury which the sample of flour has sustained. Whether, therefore, the sample contains a certain proportion of nitrogen, or whether it contains albumen, fibrine, and caseine in sufficient quantity, it may still want the very condition which is essential to the manufacture of good bread. My objection, therefore, to the mere determination, however accurate, of the proportion of nitrogen contained in wheat flour, or of the various principles which form the gluten, is, that it does not represent the value of the various samples for the only use to which they are applied, viz., the making of bread. The remarks of Mulder, the celebrated Dutch chemist, upon the subject of manures, are so applicable to this point that I cannot refrain from quoting them. "It has," he says, "become almost a regular custom to determine the value of manures by the quantity of nitrogen they yield by ultimate analysis. This method is entirely erroneous; for it is based upon the false principle, that by putrefaction all nitrogeneous substances are immediately converted into ammonia, carbonic acid, and water! But these changes sometimes require a number of years. Morphine, for example, is prepared by allowing opium to putrefy; and the process for preparing leucin, a substance which contains 10.72 of nitrogen, is to bring cheese into putrefaction. Cheese, therefore, does not perhaps in a number of years resolve itself into carbonic acid, ammonia, and water, but produces a crystalline substance, which contains no ammonia. Hence the proportion of nitrogen yielded by manures is not a proper measure of their value, and therefore this mode of estimating that value ought to be discontinued."[38] We infer, therefore, that the proportion of nitrogen furnished by food of various kinds is not the true measure of their nutritious value, and cannot for practical purposes take the place of that process by which the amount of rough gluten is determined.

No better illustration can be given of the uncertainty which attends the inferences drawn from the ultimate composition, than the fact heretofore stated in regard to hay, the nutritive value of which is placed in the tables containing the results of these analyses, at a figure nearly the same as that of ordinary wheat flour.[39] In the paper on the "Composition of Wheat," by M. Peligot—(" Comptes Rendus," February 5th, 1849)—to which I have already referred, the author gives the results of the various analyses which he has made, and details the process he adopted.

Aware of the complex and difficult nature of the examination as conducted by him, he seems to doubt in regard to some of the results given in his tables In the fourteen samples which he analysed, the proportion of water ranges from 13.2 to 15.2, which is a rather higher average than is yielded by our American samples, especially those which have not been shipped across the Atlantic. Of the nitrogenous matter, soluble and insoluble, the proportions range from 9.90 per cent, to 21.50 per cent.; the former being from a sample of very soft and white French wheat; the latter from a very hard wheat with long grains, from Northern Africa, cultivated at Verriéres. Another sample from Egypt yielded 20.60 per cent, of these nitrogenous matters, both of which are very remarkable proportions.

In describing the process for ascertaining the amount of insoluble nitrogenous matters, this author adverts to their estimation either by the quantity of nitrogen gas furnished, or of ammonia formed, the last being preferred for substances, which, like wheat, contain only a few hundredths of nitrogen. The results which he obtained by this method were compared with those yielded by the direct extraction of the gluten by softening the farina under a small stream of water. "These results," says he, "differ but little from each other when we operate upon wheat in good condition, although the gluten which we thus obtain holds some starch and fatty matter, while the starch which is carried away by the water contains also some gluten." The loss and gain, as I have already explained, and as has been proved by these and other comparisons, are nearly balanced, and the amount of rough gluten will therefore afford a fair exhibit of that of the insoluble nitrogenous matters in this grain.

The salts in the samples of wheat analysed by M. Peligot, were either wanting or were in small proportion; while the amount of fatty matter ranged from 1.00 to 1.80 and 1.90 per cent.

These results agree very well with those which I have obtained. But it is probable that the proportion is liable to great variation, inasmuch as it is inferred that the fatty matter originates from starch through its exposure to the general deoxidising influence which prevails in plants.[40] There are also many difficulties attending the accurate determination of this matter, and which are probably the cause of the higher proportion often given. It is properly remarked by M. Peligot that the ether employed in this process should be free from water, and that the flour ought also to be very dry. By neglecting these precautions, we separate not only the fatty matter, but also a certain amount of matters soluble in the water, which is furnished as well by the wheat as by the ether.

It would not, I think, be difficult to point out some incorrect views entertained by this chemist, and more especially those which relate to the fatty matter. Some of his processes for the separation of various substances, if not faulty, require so many conditions for success as to render the results, at least in other hands, exceedingly uncertain.

But the capital error which he has committed is that concerning the bran, already adverted to, which he considers injurious to the flour, chiefly in consequence of the large proportion of fatty matter which it contains.

In regard to the soluble nitrogenous matter usually called albumen, from its resemblance to the animal substance of the same name, I have to remark that in my trials the proportion has been found to be considerably less than that often given in tables of the composition of wheat. In one sample it was found to be as low as 0.15 per cant., in another it did not rise above 0.20 per cent. The amount was usually so inconsiderable, that I did not think it worth while to retard the progress of the work by following out processes which could add little to the utility of these investigations.

Although much time and labor have been expended upon the analyses of the ash of plants, I have but slight confidence in the results heretofore given. The difficulties which attend the obtaining the ash in a proper condition, and the fact that the products of all the organs and parts of the plants have been analysed together, must necessarily impair the accuracy of the experiments, and render the inferences drawn from them of uncertain value. Much, indeed I may say almost everything, still remains to be done in this department of agricultural chemistry.