dissolved in water. The powdered residue, after treatment with water, was macerated with dilute acid, and gave negative tests for alkaloids. The extractions with caustic soda for identification of albuminous substances, followed by chlorine water for the estimation of lignin and cellulose, have not yet been determined. The results of the proximate analysis as so far completed may be stated as follows:

The difference of 38.96 per cent would include pectose, coloring matter, and cellulose or woody fibre.

A qualitative ash determination showed the presence of calcium, magnesium, albuminum, potassium, sodium, a trace of iron, sulphates, phosphates, and chlorides.

Petroleum spirit extracted a solid substance, yellowish-green in color, of sp. gravity .984, melting from 84° C. to 85° C. insoluble in water, slightly soluble in boiling 95 per cent alcohol, soluble in absolute alcohol, cold ether, chloroform, amyl alcohol, benzole, carbon disulphide, oil of turpentine, and linseed oil. It was slightly acted upon by aqueous alkalies; but readily saponified with alcoholic soda. Treating the soap with lead acetate and boiling the precipitate with ether, a yellow crystalline substance was obtained, melting from 43° C. to 60° C. Sulphuric acid combined with solvents gave characteristic and distinct reaction with the yellowish-green petroleum spirit residue, and with the crystalline substance separated from it..

A scheme has been proposed for the identification of various waxes, based upon quantitative experiments.9 The examination. rendered division into two groups possible, according to the solubilities of the waxes with chloroform. Again their action with

• Contributions to the Chemistry of several varieties of wax, by E. Hirschsohn. Pharmaceutical Journal and Transactions, Vol. x, March, 1880.

ether, and acetate of lead solution added to the alcoholic solutions, allow the several varieties of waxes to be distinguished from each other.

The petroleum spirit residue was submitted to the tests proposed in Hirschsohn's scheme. It was boiled with ten times its volume of chloroform and when cool the liquid became cloudy. By this test, the petroleum spirit residue was placed in the group with Carnauba and Bahia wax. An ethereal solution of the petroleum spirit residue, on adding an equal volume of alcohol, remained clear. According to Hirschsohn's scheme, 10 an ethereal solution of Bahia wax similarly treated remains clear, and by this means, the wax is distinguished from Carnauba wax, which it is said to resemble in most of its properties. The wax from Copernicia cerifera, the Canauba tree of Brazil, and Carnauba wax obtained from the leaves of Corypha cerifera, are related very closely by their chemical properties and possibly are identical.11

Carnauba wax is described as a clear yellow wax with a greenish tinge, and harder than beeswax. It contains a notable percentage of free melissyl alcohol and other alcohols very difficult to separate. Insoluble in water, it is dissolved with difficulty by alcohol and ether, though readily soluble in carbon disulphide and oil of turpentine. It is not acted upon by linseed oil; it is changed yellow by nitric acid; with sulphuric acid no appreciable effect. The melting point is variously stated from 82° C. to 85° C. The specific gravity from .998 to .999.

A table of the specific gravity of the different kinds of waxes, prepared by Dietrich12, shows the density of animal wax to be notably low compared with vegetable waxes. Allen 13 states that the presence of vegetable wax in adulterations of beeswax is positively established if the density of the sample exceed .970.

By the method followed out in this analysis, petroleum spirit extracted from the powdered bark a substance of constant melting point, which is identified as a wax. It resembles, in its ethereal solution not clouding on addition of alcohol, Bahia wax ; in melting point and specific gravity, Carnauba wax; also the latter wax by

10 Loc. cit.

11Gmelin. Handbook of Chemistry, Vol. xviii. Translated by H. Watts, London, 12 E. Dietrich. Specific Gravity of wax. Journal of Chemical Society, 1882, Vol. xlii, p. 1139.

13A, H. Allen, Commercial Organic Analysis. (Also see in same work Tables of sp gr. waxes.)

its insolubility in water and action with nitric acid. It differs from Carnauba wax in its greater degree of solubility in absolute alcohol, ether, and aqueous alkalies. Linseed oil is an active solvent for it, but does not dissolve Carnauba wax. The color reactions of the petroleum spirit residue with sulphuric acid have been described above. It is stated that sulphuric acid produces no effect with Carnauba wax.14

The wax obtained from the bark of Fouquieria splendens differs generally in its properties from known vegetable waxes, and is evidently a new wax, peculiar to this plant. I propose that it be called Ocotilla wax.

In the ether, absolute alcohol, and water extracts, the presence of an acid resin, a white crystalline substance, gum resin, glucose, possibly glucosides, gum, and a red coloring matter were indicated.

The investigations described in the preceding pages were conducted in the Chemical Laboratory of the Philadelphia College of Pharmacy, August and September, 1884.

THE DENSITY OF SOLID CARBONIC ACID. By Prof. JAMES Dewar, Cambridge, England.

[ABSTRACT.]

THE object of this paper is to compare the theoretical density of the solid as deduced from the formulæ of Clausius, with experimental number found by the author.

By

THE COMPOSITION AND METHODS OF ANALYSIS OF HUMAN MILK1. Prof. ALBERT R. LEEDS, Stevens Institute, Hoboken, N. J. In a previous paper, "On Infant Foods," which I read before the College of Physicians of Philadelphia, May 2, 1882, I alluded briefly to the investigation then in progress upon the composition of human milk, and gave a tabular statement of the minimum, maximum, and average results of the analysis of forty-three samples, the total number of analyses which I had made up to that date.

14A. B. Prescott. Outlines of Proximate Organic Analysis.

The tables and chart are from electrotypes of advance sheets from the Transac tions of the College of Physicians of Philadelphia.

Since then I have analyzed thirty-seven more samples, and have verified the results by a critical examination of the various methods of analysis which are in use at the present time.

In the beginning of my previous paper, I asked the question, "What is human milk?" and stated that to answer this question satisfactorily, we should know at least three things: 1st, All the constituents; 2d, Their relative proportion; 3d, Their chemical and physiological properties. It is not my present object to discuss the first and third points, except in the light of the following isolated results which were only incidental to the main object of the present inquiry.

The albuminoids and fat of a large number of samples, as obtained by precipitation with Ritthausen's solution, were extracted with ether, until the albuminate of copper ceased to give up any further traces of fat to the solvent. It then became after drying a very light-green amorphous powder.

In order to separate the albuminoids, this powder was digested with very dilute hydrochloric acid, which carries some of the organic matter into solution along with the copper. The residue, after washing and drying at 100°, formed brownish, somewhat brittle, amorphous masses.

The percentage of cupric oxide contained in this albuminate of copper was found in two analyses to be 20.93 per cent and 20.63 per cent. The ultimate analyses of the albuminate, after deducting cupric oxide, yielded the following results:

Two analyses of the total albuminoids, left behind after the foregoing treatment of the albuminate of copper with dilute hydrochloric acid, gave the following figures:

The albuminoids, separated by hydrochloric acid from the copper albuminate, were digested with 50 per cent alcohol, at the boilingpoint. On cooling, the filtered solution deposited a white, flocculent,

voluminous precipitate, whilst the residue on the filter formed a somewhat brownish mass.

This precipitate would correspond to the "caseo-protalbin" of Danilewsky (Jahresb. der Thierch. 1880, 186), and the residue, which was much the larger in amount, to his "caseo-albumen.”

An analysis of the caseo-protalbin gave :

An analysis of the "caseine," obtained by Makris in quite a different manner from woman's milk, yielded for its composition, carbon 52.35 per cent; hydrogen 7.27 per cent; and nitrogen 14.65 per cent.

The foregoing analyses render evident that the bodies examined are not homogeneous, and in every case the process of separation left behind mixtures of substances the true nature and composition of which are at present unknown. Moreover, the deportment and properties of the bodies examined are such as to lead one to the conclusion that the investigation was being conducted, not upon bodies in the condition and with the properties which they possessed originally in woman's milk, but upon substances whose composition and properties had been altered by the operation of the reagents employed.

It had been the intention of the author to endeavor to isolate the various constituents, at present very imperfectly known, of the fat of human milk. This desire had been increased by the fact that the ethereal extract of the copper albuminate, obtained from numerous samples, although not by any means from all, was colored emerald-green by some copper salt. As to the chemical nature of this copper salt, going as it does in perfect solution in ether, I have no knowledge. Unfortunately, the entire mass of fats was lost by accident in the early stages of manipulation, and I shall have considerable difficulty in again procuring sufficient material to work upon. No sample of cow's milk which I have analyzed yielded to ether an emerald-colored solution. The unknown body is peculiar to woman's milk.

METHODS OF ANALYSIS.

Passing by the methods of analyses which were employed when the nature of the difficulties to be overcome was imperfectly under