the apparatus, it is heated, exhausted by means of a Sprengel air-pump, cooled and weighed. The portion containing the copper oxide is then heated in an appropriate air-bath and a gram or more of hydrogen is passed in. The hydrogen is converted into water by the copper oxide and the water formed condenses in the side-tube referred to above. The apparatus is then cooled and weighed. After weighing, any gases which may remain in the apparatus are pumped out and analyzed, and thus a correction which may be necessary for impurities in the hydrogen is obtained. The apparatus is heated at the same time and all of the water formed is expelled. By weighing again, the amount of oxygen taken from the apparatus is determined.

The advantages of the method are: the weight of the hydrogen is found directly as the difference of two weights subject only to a small correction for impurities; the weight of the oxygen is also found directly and independently of the weight of the hydrogen; each is weighed in a vac. uum, and by using a proper counterpoise any correction for the buoyancy of the air becomes unnecessary; impurities in the hydrogen and especially any nitrogen which it may contain will be detected; finally, there can be no error from incomplete combustion of the hydrogen.

HAMPE'S METhod of determining Cu2O IN METALLIC COPPER. By Prof. FRED. P. DEWEY, Smithsonian Institution, Washington, D. C.

[ABSTRACT.1]

=

THE author found that Hampe's reaction 3Cu20+6 AgNO3 + 3H2O H6Cu4N2012 +2CuN2O6 +6Ag, whereby two-thirds of the copper is converted into an insoluble nitrate and remains with the Ag precipitated on treating pure Cu2O, with AgNO, in neutral solution, is modified by the presence of Cu as follows:

4

2 4

4Cu2O + Cu + 6AgNO3 + 6H2O=2H¿Cu ̧N2O12 +CuN2O、, + 6Ag. That is, all the Cu2O present in the metallic copper remains in the insoluble material.

ON THE CONSTITUENTS OF WINTERGREEN LEAVES. (GAULTHERIA PROCUMBENS, LIN.) By Prof. FREDERICK B. POWER and NORBERT C. WERBKE, Madison, Wis.

[ABSTRACT.]

IN this investigation the authors have isolated the hydrocarbon or terpene from the volatile oil of wintergreen, which was stated many years ago by Cahours (Ann. Chim. Phys. (3), 10, p. 358) to be contained therein,

1 Published in full in Proc. U. S. National Museum for 1888.

and have confirmed and determined some of its physical properties. They find the terpene to be present in very much smaller amounts than has been previously stated and generally accepted. The distinguishing characteristics of the oil of wintergreen and the closely related volatile oil of sweet birchbark (Betula lenta, Lin.) are also demonstrated.

Experiments were furthermore made to ascertain whether the crystalline poisonous principle, isolated by Professor Plugge, of Groningen in the Netherlands, from the leaves of several ericaceous plants, and termed by him andromedotoxin, is also contained in the leaves of the Gaultheria or wintergreen. This is proved not to be the case.

ON THE OXIDATION OF NITRO-P-XYLENE WITH POTASSIUM FERRICYANID E By Prof. W. A. NOYES, Rose Polytechnic Institute, Terre Haute, Ind.

[ABSTRACT.]

A SERIES of experiments has been conducted for the purpose of discovering any general laws which may govern the oxidation of benzene derivatives by potassium ferricyanide. One of the most interesting results of the work has been the discovery, by Moses and myself,' that meta-nitrotoluene is oxidized with much greater difficulty than ortho-or para-nitrotoluene. This rendered it probable that if mono-nitro-para-xylene were subjected to the action of the agent in question, it would yield ẞ-nitro paH 1. ra-toluic acid 2. a substance unknown at the time this CH 4.. investigation was begun. On carrying out the oxidation, the acid was obtained, as expected, but only in small quantities. Larger amounts of the acid were obtained by starting with nitro-para-toluidine and converting it by means of Sandmeyer's reaction. This method has recently been used independently by St. Niementowski."

Barium, calcium and copper salts of both the nitro and the amido acid were prepared and analyzed.

A second product of the oxidation was a substance having the composition of nitro-tere-phthalic acid with the addition of two molecules of water. As this substance could not be obtained by crystallizing nitro-terephthalic acid under a variety of conditions, and was only formed by the oxidation of nitro-xylene by potassium ferricyanide, the water is probably water of constitution and not water of crystallization. The substance C(OH)3 probably has the formula, Ce H3 NO2, and its formation is analogous C(OH),

to the formation of ortho-sulphamine acids where sulphinides would be expected.3

1 Am. Chem. Journal, 7, 149.

2 Ber. d. Chem. Ges., 21; 1992. Am. Chem. Journal, 8, 179.

PROPYLIDENE DI-ETHYL AND DI-METHYL ETHERS. By Prof. SPENCER B. NEWBURY, Cornell University, Ithaca, N. Y.

[ABSTRACT.']

By the action of ethyl alcohol and methyl alcohol on propionic aldehyde in presence of glacial acetic acid, propylidene di-ethyl ether and propylidene di-methyl ether were obtained, the former the next higher homologue of ordinary acetal. The reactions which took place are as follows:CH3–CH,–CHO+2C,H,OH=CH3–CH,−CH(OC,H,),+H,0. CH ̧—CH,—CHO+2CH3OH = CH ̧—CH¿—CH (OCH3)2 + H2O. Propylidene di-ethyl ether boils at 122°C.; spec. grav., 0.8473 at 0°. Propylidene di-methyl ether boils at 86°-88°C.; spec. grav., 0.8657 at 0°.

A NEW VEGETABLE DYE. By Dr. THOMAS TAYLOR, U. S. Department Agriculture, Washington, D. C.

[ABSTRACT.]

A RICH crimson liquid extracted from a mushroom of the genus Lactarius, when applied to paper produces a delicate salmon tint. An alkali applied to the paper thus stained changes the tint to lilac, thus affording a test for alkalies. The lilac color changes again to salmon on the application of a weak acid.

FATAL POISONING BY CARBON MONOXIDE. By Prof. W. P. MASON, Troy, N. Y.

[ABSTRACT.]

THREE deaths, and a number of serious illnesses, were caused by the breaking of a "fuel gas" main, and the consequent escape into adjoining houses of an odorless gas containing 37.5 per cent carbon monoxide. A specimen of blood which was taken from the heart cavity of one of the victims, although twenty months had elapsed since it was collected, still presented the same brilliant red color, and gave the spectrum absorption bands as strongly as when fresh.

THE SAFETY OF COMMERCIAL KEROSENE OILS. By Prof. SPENCER B. NEWBURY, Cornell University, Ithaca, N. Y.

[ABSTRACT.']

TESTS made by the author show that the oil in modern lamps with circular or double wick often becomes heated to 110° to 112° F., a tempera

1 The full text of this paper will appear in the American Chemical Journal.

ture considerably above the usual legal flashing point. As, however, it is not yet proved that the vapor of oils of high flashing point is capable of giving dangerous explosions with air at any temperature, experiments were made in the hope of establishing a clear relation between the flashing point and the danger of explosion.

The following conclusions were reached:

1. All the paraffines up to decane, and probably also higher members of the series, will at suitable temperatures form explosive mixtures with air.

2. Oils which flash at a point considerably above 100° F. may under proper circumstances give violent explosions at their flashing temperature.

10

3. An oil consisting of pure decane, C10H22, would be accounted a safe oil by the legal flashing point test of New York state, while one consisting of nonaue, C,H20, would be below the standard.

A sample of illuminating oil of good quality yielded on fractional distillation about one and one-half per cent of octane, and notable quantities of nonane and decane. To samples of the same oil, and also to samples of" mineral sperm " oil, varying proportions of pentane, hexane and heptane were added, and the reduction of the flashing points noted.

These experiments show that an oil heated above its flashing point is dangerous, whether that point be high or low. Oils which flash below 120° F. are not safe to use in the powerfully-heating lamps which are now so largely employed.

LAKE ERIE WATER AT CLEVELAND, OHIO. By ALBERT W. SMITH, Cleveland, Ohio.

[ABSTRACT.]

AN account of a chemical examination of the water with reference to locating the new tunnel of the Water Works' Department.

SOME NOTES ON PROGRESS IN CHEMICAL METHODS OF WATER ANALYSIS WITH ESPECIAL REFERENCE TO THE AMMONIA PROCESS. By Prof. FRANK H. MORGAN, Cornell University, Ithaca, N. Y.

[ABSTRACT.]

THE lines of enquiry which now demand the attention of workers in this direction may be summarized as follows:

First. A careful revision and adoption of uniform methods of procedure for those processes at present in general use.

Second. Careful investigation of modifications of the "ammonia" and

"oxygen" processes, designed to render more definite results as to the nature and amount of organic matter in a water.

Third. Continued studies of the conditions of nitrification in the soil and its relations to purification of both well and river waters.

Fourth. Trials of new reagents in a search for means of distinguishing different kinds or conditions of organic matter.'

Fifth. A careful study of the relations of time, temperature, concentration and pressure to the action of reagents on the organic matter.

Sixth. A complete index to the literature of potable water analysis and carefully drawn monographs of each method in general use.2

The following lines of investigation are being carried on by the writer: 1. Trial of the action of a constant volume of concentrated alkaline permanganate.

Results so far obtained show no decided increase in the amount of nitrogenous matter thus converted into ammonia.

2. Trial of the amount of ammonia that may be obtained by a preliminary treatment with H2SO4. Application of the principles of Kjeldahl's process was suggested by Prof. E. Waller, Jour. Am. Ch. Soc., 8, 150, but no results have as yet been published.

Present progress of the writer shows the possibility of a considerable increase in the amount of ammonia as compared with the total ammonia obtained by the ordinary process, but satisfactory regularity has not yet been secured.

3. Determination of the amount of ammonia developed by the action of Na2CO3 on the organic matter during the ordinary distillation of free ammonia.

The method used is to determine the real saline or free ammonia by direct test and subtract this from that obtained by the ordinary distillation with Na2CO3.

The following results have been obtained:

A well water from a suspected district (alb. am. 0.10 and ni trates n=5. p. p. m)..

Many experiments have been made upon the details of the ammonia process. A small amount of CO, in a distillate may diminish the color developed by Nessler reagent. The amount of Nessler to be added to the

'Practically only one reagent has been in general use.

2 The writer is making an index to the "Ammonia Process" and hopes to present a monograph at the next meeting.