of a thin section of this variety of eosphorite showed the quartz scattered as grains through the mass.

Deducting the residue from the analysis and calculating again to 100 61, we have

This is very nearly the composition of eosphorite as analyzed by Penfield. The excess of lime is in part due to the presence of apatite which is found associated with much of the eosphorite. The compact mineral is simply eosphorite intimately mixed with quartz and other species found at the locality. The greener colored varieties contain dickinsonite and are somewhat more fusible than pure eosphorite, while the lighter colored varieties such as analyzed are more difficultly fusible. The density of these varieties varied from 2.92-3.08.

The name eosphorite is from the Greek wo pópos (a synonym of poopópos, whence the name phosphorus), which means dawn-bearing, in allusion to the characteristic pink color of the crystallized mineral.

2. TRIPLOIDITE.

Physical characters.-Triploidite occurs in crystalline aggregates which are distinctly parallel-fibrous to columnar in some cases, and in others divergent; and again confusedly fibrous to nearly compact massive. Occasionally individual prismatic crystals are distinct, being separated from one another by the transparent quartz in which they are imbedded and from which they become detached when the mass is broken into small fragments. The isolated crystals have sometimes a length of an inch or more, but it is not possible to detach them except in very small pieces. The conditions are obviously extremely unfavorable to the formation of terminated crystals, but a careful and long-continued search upon a large amount of material was at last rewarded by the discovery of a few more or less perfect specimens. In rare instances the crystals have been observed standing free in small cavities in the massive mineral. The crystals have perfect orthodiagonal cleavage.

The hardness of triploidite is 45-5, and the specific gravity 3.697. The luster is vitreous to greasy-adamantine. The color is yellowish to reddish-brown, in the distinct crystals also topaz to wine-yellow, and occasionally hyacinth-red. The streak is nearly white. Transparent to translucent. The fracture is subconchoidal.

4.

Crystalline form.-Of the few terminated crystals obtained, three only were suitable for measurement and only one of these had the terminations complete. These were extremely small, but the planes were of so high a luster that they gave good reflections, but little inferior to those obtained from the best eosphorite crystals. The planes in the prismatic zone are in the larger crystals so much striated as to admit of no satisfactory measurements. In the crystals selected for careful measurement the only planes in this zone which could not be used at all were the clinopinacoids, for the others the reflections were reasonably good. The crystals show occasionally false planes, bearing no relation to the axes of the crystal, and which are evidently impressions of portions of adjoining crystals.

a

The crystals belong to the MONOCLINIC SYSTEM and their habit is shown in figure 4. The axial ratio was obtained from the following fundamental angles :

These angles are good, though a little less so than those given for eosphorite-the probable error, however, does not exceed +1'. The axial ratio is:

The following are the principal angles, both calculated from the above data, and as measured on the same crystal (1) and on the two others (2 and 3):

A comparison of the above angles with those given by Brooke and Miller for wagnerite shows that the two species are homoeomorphous.

Thus in the three diametral zones, we have :

As the crystal of wagnerite is placed by Miller, the planes g, a, cand e have the symbols (120), (100), (001), (021) respectively. In the figure given by Miller the prism g 120 (= I, (110) triploidite) has the greatest development; it was made the unit prism by Naumann.

Optical properties.-The only point that could be established in regard to the optical character of triploidite was the position of the axes of elasticity. The crystal used for measurement had the clinopinacoid so far developed that it could be examined directly in a Rosenbusch microscope. It was found that of the two axes which lie in the plane of symmetry, one very nearly coincides with the vertical axis, being inclined behind (see fig. 4) 3°-4°, and the other consequently is almost normal to the orthopinacoid. The position of the optic axes could not be fixed. The crystals show no perceptible absorption phenomena.

Chemical composition.-Triploidite was analyzed by Mr. Penfield. This hydrous phosphate was found to contain iron and manganese, both being in the lowest state of oxidation, with a small amount of lime; it is entirely free from fluorine. The method of analysis was substantially the same as that of eosphorite, (described on page 39). There being no alumina in the mineral, the phosphoric acid was determined directly from the solution of the fusion. The fusion did not effect a complete separation of the phosphoric acid from the iron and manganese, as it was retained by the small amount of lime present. It was weighed with the iron, and afterwards was separated from the iron by means of ammonium molyb

date, determined, deducted from the weight of the latter and added to the phosphoric acid determination. The results of two analyses are:

2 5

The ratio P,O,: RO: H,O=1:4:1 corresponds to the formula R.PO,+HỌ, or R‚P,O,+H,RO,, where R=Mn: Fe =

2

2 8

3:1. This formula requires:

2

21

Among the other phosphates and arsenates the following seem to be closely related to triploidite in composition:

None of these species has any relation to triploidite in crystalline form. On the other hand the similarity between the angles of wagnerite and triploidite has already been shown; moreover, the composition of triplite is analogous to that of wagnerite and for these reasons a relation between triplite and triploidite immediately suggests itself. The composition of these minerals is:

Wagnerite
Triplite
Triploidite

Mg3P2O8+MgF,

(Fe, Mn),P,O,+ (Fe, Mn) F2
(Mn, Fe), P20, +(Mn, Fe) (OH),

It should be stated that the perfect transparency and brilliant luster of the crystals analyzed prove beyond all question that the absence of fluorine and the presence of water (determined directly) are not due to any alteration. The fact that all the bases are in the lower state of oxidation would be confirmatory evidence were it needed. The conclusion to which we are led is this-that in the compound triploidite the radical hydroxyl (OH) plays the same part as the element fluorine, the molecule R(OH), taking the place of the RF2.

Pyrognostics. In the closed tube triploidite gives neutral water, turns black and becomes magnetic. Fuses quietly in the naked lamp flame and B. B. in the forceps, colors the flame green. Dissolves in the fluxes, giving reactions for manganese and iron. Soluble in acids.

An analysis of another specimen of triploidite gave P,O, 32-24, FeO 18-65, MnO 42.96, CaO not determined, H,O 4·09, quartz 109. The lime was accidentally lost but calculating from the amount of phosphoric acid retained by the iron it amounted to 0.90 per cent. The analysis is interesting as showing that the iron and manganese vary in different specimens, the darker colored varieties containing the most iron.

The name triploidite given to this species, from triplite, and edos form, indicates its resemblance to triplite in physical characters, and its relation in chemical composition.

(To be continued.)

ART. V.-On Dinitroparadibrombenzols and their Derivatives; by Dr. P. TOWNSEND AUSTEN, F.C.S., Assistant Prof. of Chemistry in Rutgers College. Third paper.*

IN my former papers, I have described the formation of three dinitroparadibrombenzols, and proved the a and ẞ variations to be isomeric compounds. With regard to the third, I am still somewhat in doubt.

The peculiar formation of nitroparadibromaniline by treatment of alpha dinitroparadibrombenzol with ammonia, has led me to make experiments with other reagents, and I have been gratified at encountering some quite unexpected phenomena. These I shall mention in another paper.

Beta-dinitroparabromphenol.

By pouring a very concentrated alcoholic solution of potassa over the beta-dinitrodibrombenzol, the mass became scarlet-red, indicating the formation of a salt. Examination showed, however, that much of the substance was left unaffected. On heating, an action set in, and fine bubbles were formed. On diluting with alcohol and acidifying with hydrochloric acid, a dark brown flocculent mass was obtained, insoluble to any extent in alcohol. It was soluble in glacial, acetic acid and acetic ether, separating in the form of an amorphous powder. As it was also soluble in a solution of sodium hydrate, and was precipitated therefrom by hydrochloric acid, I take it to be an azoxyphenol.

Various attempts to obtain a good yield of the phenol by direct treatment with potassa, or soda, in different amounts and solvents, did not meet with success, except on a small scale. Although in most cases, the phenol salt was formed, as could be discerned from the red color of the liquid, yet on application * Compare this Journal, III, ix, 118, and xiii, 95.