of carbon is quadrivalent toward oxygen at the ordinary temperature and under ordinary conditions. How otherwise shall we explain the formation of carbon dioxide in the processes of decay, fermentation, etc.? But the atom of carbon is just as positively quadrivalent at high temperatures.

The comparative ease with which carbon monoxide takes up chlorine appears to prove that it possesses free affinities. But if we accept this as a proof of the existence of free affinities in carbon monoxide, we have still better grounds for believing that free affinities are present in ethylene, for this gas combines with chlorine much more readily than carbon monoxide does. Still the view is commonly held that in ethylene the two carbon-atoms of the molecule are united by the mutual action of two affinities of each atom.

These considerations show that the nature of carbon monoxide is, as yet, but very unsatisfactorily understood. The first question which suggests itself is this: How far are we justified in considering carbon monoxide as a body possessing free affinities?

If we attempt to answer this question entirely without prejudice, we see that the principal experiment which is supposed to prove the existence of free affinities in carbon monoxide is the above mentioned experiment with chlorine. Oxygen does not combine with carbon monoxide at the ordinary temperature. This is readily understood, for, in order that the carbon monoxide and oxygen may combine by direct contact of the two substances, the oxygen-molecule must first be decomposed into its constituent atoms. An interesting experiment in this connec tion has been described by E. Ludwig, who shows that carbon monoxide is oxidized by chromic acid at the ordinary temperature forming carbon dioxide. In this case carbon monoxide is active enough to separate one atom of oxygen from chromic acid and to employ it for the formation of carbon dioxide.

We have occupied ourselves with an experiment similar to that described by Ludwig, and have obtained a different and unexpected result. It appeared to us to be of interest to know whether, at the ordinary temperature, ozone has the power to transform carbon monoxide into the higher oxide. According to the views which are commonly held concerning the nature of the substances experimented upon, the transformation mentioned could be predicted with a tolerable degree of certainty. Particularly is this the case, if we consider the result of Ludwig's experiment, for usually ozone gives up its extra atom of oxygen with still greater readiness than chromic acid does. There is indeed no substance in the whole field of chemistry which furnishes us with a better means for obtaining

* Annalen der Ch. u. Pharm., clxii, 47.

a free atom of oxygen than ozone. If then we bring in contact with ozone a substance, which in turn is capable of taking up an atom of oxygen without itself undergoing change; which, indeed, possesses an attraction for oxygen, we are certainly justified in expecting to see the two substances act upon each other. But the experiment gave the unexpected result that ozone does not act upon carbon monoxide.

Two very careful experiments were performed. Pure carbon monoxide free of dioxide was first collected in a gasometer. This was then conducted from one side through three cylinders containing potassic hydroxide and lime-water into a flask. From the other side a current of oxygen was conducted through potassic hydroxide and lime-water, and then through a tube, in which the oxygen was converted into ozone, into the same flask. This flask was provided with a stopper having three holes. From the third hole a tube led to a cylinder containing lime-water; and this cylinder was connected with a final cylinder containing potassic hydroxide. Let us see what purposes the different parts of the somewhat complicated apparatus served. In the first place, the carbon monoxide was caused to pass through potassic hydroxide and lime-water in order to absorb every trace of carbon dioxide which might be present. The oxygen was treated similarly for a similar purpose. The ozone generator employed was that described by Wright* for use with the Holtz electrical machine, the best conditious being retained throughout the experiment for the working of the apparatus. The pure carbon monoxide and the ozonized oxygen were then caused to meet in the final flask, the inside of which was moist, as, for some unknown reason, ozone does not exhibit its oxidizing properties as well when dry as when moist. The mixture of the two gases, and any carbon dioxide which might have been formed, were then passed together into lime-water, contained in a cylinder, the lime-water being protected from the influence of the carbon dioxide of the air by the potassic hydroxide contained in the last cylinder.

Slow currents of carbon monoxide and oxygen were now passed through the apparatus, and, although the action was continued for a long time, not a trace of a precipitate could be detected in the last cylinder, containing lime-water. The strength of the gas-currents was frequently changed, but nothing brought about the expected result.

In view of the importance of the experiment we were not satisfied with this one form of it. As direct sun-light greatly facilitates the combination of carbon monoxide with chlorine, it seemed probable that it would be of service in causing the combination of the two gases under examination; and, accord

* This Journal, vol. iv, July, 1872.

ingly, we repeated the described experiment with the following modifications: The final flask, above mentioned, in which the carbon monoxide and the ozone were brought together, was replaced by two large glass balloons, and these were placed in the direct light of the sun. Again slow currents of carbon monoxide and ozone were passed through the apparatus for hours, the rapidity of the currents being varied at different times.

In this case also we obtained only a negative result. We hence are in a position to assert positively that carbon monoxide is not oxidized by ozone.

If we now bear in mind that ozone acts destructively upon a great many saturated stable compounds, that one of the atoms of the ozone molecule has a great tendency to unite with other bodies, then the result of the above described experiments remains inexplicable. It shows at all events that carbonmonoxide itself, at the ordinary temperature, has no very great tendency to unite with oxygen, for, if our ideas in regard to the nature of ozone are correct, the conditions for such union were very favorable in our experiment.

We hope gradually to be able to experiment more fully upon this interesting subject with the object of collecting material which may enable us better to understand the nature of the so-called non-saturated compounds. We propose next to study the action of hydrogen peroxide upon carbon monoxide. December, 1875.

ART. XVI.-Mineralogical Notes; by EDWARD S. DANA.-No. I. On the Optical Character of the Chondrodite of the Tilly Foster Mine, Brewster, New York.

*

IN a memoir on the Brewster chondrodite, published in the third volume of the Transactions of the Connecticut Academy, I have given the results of an optical examination of chondrodite crystals of the second type. It was there shown that the optic axes lie not in the basal plane, but in a plane making an angle of about 154° 10' with the base; and, in consequence, that the crystals of this type, at least from that locality, belong optically not to the orthorhombic system, but to the monoclinic, while the various measurements proved that the deviation in angle from the orthorhombic type could not be greater than 2 or 3 minutes. A recent repetition of the measurements with the stauroscope on the same crystals, and also on another not examined before, confirm the results ob

*See also this Journal, III, ix, 63, for extracts from the paper.

tained, and leave no room for doubt on the subject. The following is the evidence on this point thus far obtained.

Measurements on four independent crystals gave for the supplement angle made by the plane of the axes:

I.

With e1(203), 18° 9'; hence with basal plane, 25° 50'.
With e2(2=201), 45° 9';

66

66

25° 46'.

66

66

66

66

66

66

66

66

II. With ea(205), 40° 55;
III. With B (i-i-100), 65°-70°;

IV. With the basal plane, direct measurement,

25° 59'. 20°-25°. 25°.

I have since made an optical examination of a crystal of the third type. One single crystal of this type allowed of a stauroscopic examination. Only a small portion of it was transparent enough for use, but the circumstances allowed of a very exact adjustment according to the method of Groth, and the probable error cannot exceed one degree. The measurement gave for the supplement angle between the base and the plane of two of the axes of elasticity 74°, a result which, like the corresponding one obtained for the second type, is at variance with the supposed orthorhombic character of the species. The series of measurements were made at different times with independent adjustments, but no considerable variation was found in the result, so that it may be considered as being above question. It is remarkable that the correspondence between the two types is- not greater. In crystalline form the third type is between the first and second. I have to regret that no satisfactory material is at hand for the extension of these investigations to the Vesuvian humite.

It may not be out of place to state here that, through the kindness of Mr. Cosgriff, the Yale College Cabinet has recently received some exceptionally large crystals of chondrodite from the Tilly-Foster Iron Mine. The crystals were quite perfect, and four inches or more in length. Like all the large crystals they are partially altered, and have therefore little luster. They are penetrated with serpentine and brucite derived from their alteration.

ART. XVII.-On Hermannolite, a new species of the Columbium group; by CHARLES UPHAM SHEPARD, Sr., Mass. Professor of Natural History in Amherst College.

IN vol. 1, p. 90, of this Journal (1870), I described as probably new, a Columbium mineral from Haddam, Connecticut, to which in June last* I gave the name of Hermannolite, in honor * See Popular Guide to the Museums of Amherst College, p. 71.

of Dr. R. Hermann of Moscow, to whom chemistry has been so much indebted for the elucidation of this difficult group of minerals. By reference to my description of the mineral it will be seen that I went no further than to determine the proportions. of the bases, and of the metallic acids with which they were united, without attempting to ascertain the order in which the latter were present. I thus found:

Metallic acids,
Protoxide of iron,

Protoxide of manganese,

78.30

13.86

7.72

100.28

Desirous of learning the exact proportions of the different acids, I availed myself of an opportunity during the past summer of sending specimens to Mr. Hermann for this purpose. He has had the goodness to perform the analysis, and to com. municate to me his results in the following letter.

"Your opinion that the mineral from Haddam, which you most kindly named for me, was not columbite has been fully corroborated: for it contains no hyponiobous acid (Nb2O3), as the columbite does; but niobous acid (NbO2); and, in addition, hypoilmenic acid (IO), and also, a small quantity of hypotantalic acid (Ta2O5). The chemical formala is therefore quite different from that of the Columbite: i. e., not RO, Me'0', but 2(2RO, 3NbO2)+(RO, Me2O5)Il2O3=(‡Ta2O3+3Il203). The result of the analysis was:

Oxygen.

Calculated.

The lower specific gravity of the mineral observed by you as well as the easy solubility in sulphuric acid of the metallic acid present, are readily explained from their small content of tantalic acid, and from the greater proportion of oxygen in the niobous acid as compared with that of the hyponiobous acid in Columbite."

Moscow, Nov. 9, 1875.

The physical characters of the mineral are given in the volume of this Journal above referred to.