mon; and in some cases the exposed exterior of the pure mineral has taken a chocolate-brown color.

As has been stated, indistinct crystals and crystalline masses are not uncommon, and some few specimens, especially those placed in my hands by Mr. Ralston, admitted of exact determination. The general habit, and the more common of the occurring planes are shown in figure 1, the additional planes of figure 2 are rather rare.

The characteristic feature of all the crystals, almost without exception, is the nearly right-angled edge between the macrodomes, 1-7. Not infrequently the elongation of this terminal edge gives the crystals a prismatic appearance in that direction. More generally, however, the crystals are elongated vertically in the direction of the prism as taken in the figures; a radiated arrangement in the groupings of the crystals, sometimes observed, is a feature deserving mention. The prismatic planes (I and i-2) are in all cases narrow so that the general habit is that of a rectangular prisın; frequently the crystals are flattened in the direction of the brachypinacoid it, and upon the surface of this plane are sometimes observed a number of small outlined crystals similar to those often occurring on the diametral planes of columbite. Occasionally, also, i-i is the more prominent, giving rise to forms flattened in this direction.

The occurring planes, as seen in the figures, are as follows: i-i, i-i, I, i-2, 1-7, 1, 3-. The planes are without luster and often quite rough, so that approximate measurements alone were possible; and in different crystals some of these angles varied considerably. The angles obtained from the best formed crys tals are as follows:

^

i-2 i-295°; 1-7 1-7 = 93°. From these measurements the following axial ratios are obtained:

c' (vert.) 0.949, ō (macr.) 1.833, ă (brach.) 1.000.

Some of the calculated angles for the other forms are as follows, the angles obtained by measurement being given in parentheses:

I^ I=122° 46', I^i-i=151° 23′ (152°), i-1= 110° 35' (110°), i-i1=130° 7', / 1=137° 14', i-i3 125°55′(126°), i-73= 135° 46′ (136°).

The position here adopted shows most favorably the probable relation of the samarskite, of North Carolina, to the crystals of yttrotantalite described by Nordenskiöld.

The parentheses indicate that the forms referred to have not been observed. The pyramidal planes of samarskite (1 and 3-) are not known on yttrotantalite. It will be observed that while in the prismatic zone the agreement between the two species is close, in the domes the variation is considerable.

The prism of samarskite referred to the axes of columbite (Dana's Min., p. 516) is i-, and on this basis the other planes become as follows: i-2=i-, 1=1-3, 3-7=2, 1-7=1-ī.

The crystalline form of euxenite has not been very clearly made out, but it seems to be closely related to that of samarskite; for II, Dahl gives 126°, Greg 120°? (122° 46' samarskite); -im-i-154 and 153°, but -727-152° 13′ (samarskite); also two pyramids are mentioned giving the angles i-ip=107 (i-ž^1=110° 35′ samarskite), i-ī^p2=136° (135° 46' samarskite).

The method of association of crystals of samarskite and columbite at Miask (to be mentioned later) seems to suggest that the broad plane, - of the figure, may possibly correspond to the plane ii of columbite. (To avoid confusion it must be noticed that i-i columbite, Dana's Mineralogy = i- Naumann, and / Dana = i-3 Naumann.) This idea is supported by a single one of the specimens under examination, where of two associated crystals, the cleavage plane (probably i-i) of the columbite was exactly parallel with the plane of the samarskite called i-i above. If now this change is made, the planes, before mentioned, become as follows: If i-2=1 and i-ii-i then I=i-2, 1=1-2, 3-5 =2-3. The consideration of all the facts, however, seems to show that the method first proposed should be adopted.

It may also be mentioned here that several of the minerals of this group show angles of 91°-95°, 128, etc., in the prismatic zone, although in the other zones there is no apparent correspondence, and the habit is quite different.

The occurrence of two other minerals of this tantalic group has already been mentioned. One of these minerals occurs in regular octahedrons, sometimes nearly an inch across, with the

cubic planes, and also the form 3-3. It has a yellowish-brown color and resinous luster. Professor Brush reports, from his examination, that in blowpipe characters it agrees closely with pyrochlore; but its specific gravity as determined by him on a pure crystal is 4-794, which is considerably higher than that of pyrochlore (4-203, Hermann), so that it may approach more nearly to microlite. For a definite knowledge of its character we must consequently wait for the chemical analysis which Professor Allen proposes soon to undertake. These octahedrons occur generally in a rusty gangue, the mass of which seems to consist mostly of the same mineral. They are also sometimes observed implanted directly upon the samarskite.

The second associated mineral is columbite. It occurs in crystalline masses of considerable size, imbedded in the samarskite, or implanted upon it. The form where distinct is very similar to those given in Dana's Mineralogy, figures 429, 430, p. 516, and the angles agree closely. From some qualitative experiments Professor Allen finds that it contains a considerable quantity of tantalic acid. On this account it is a matter of some surprise that its specific gravity is only 5476.

This intimate association of columbite and samarskite at this locality is the more interesting in that, as long ago shown by Hermann, these two species occur together at Miask in the Urals. Some Uralian specimens recently examined by me have the minute crystals of columbite, well formed, implanted on the samarskite, the crystals of the two appearing to occupy a parallel position. It would here hardly be suspected that the two minerals were distinct, except from the cross fracture, in which the two decidedly differ. The American specimens, on the other hand, with the single exception alluded to, show no relation at all in the position of the crystals of the two species.

Professor Allen is at present engaged in a thorough chemical investigation of the various minerals, which have been mentioned, and the results of his work will be awaited with much. interest.

ART. XXVII.-The Effect of Silicic Acid upon the Estimation of Phosphoric Acid by Ammonium Molybdate; by E. H. JENKINS.

THE idea seems to be general that the presence of silicic acid in solutions, impairs the accuracy of the estimation of phosphoric acid by the molybdic method. In Rose's Handbuch der Analytischen Chemie, 6th edition, volume ii, page 519, under a description of this method the fact is stated that silicic acid gives a precipitate similar to the ammonium

phospho-molybdate, and Fresenius (Quantitative Analyse, 5th edition, page 334) advises the separation of silicic acid as a preliminary. W. Knop has observed (Chemisches Centralblatt, 1857, page 691) that ammonium molybdate, added to a solution containing silicic acid and a large quantity of ammonium chloride, produces a lemon yellow precipitate much like ammonium phosphomolybdate. Without this excess of ammonium chloride no such precipitate forms, either in the cold or after heating to boiling.

It might readily be supposed, however, that, though not precipitated by itself with ammonium molybdate, silicic acid could come down in a precipitate of ammonium phosphomolybdate and introduce an error. To ascertain whether this actually happens the following experiments have been made. A solution of potassium silicate was employed, made by boiling pure silicic acid with potassium hydrate, and acidifying slightly with nitric acid. It contained in 50 c.c. 2055 grs. of silicic acid, and bare traces of phosphoric acid. A solution of pure hydrodisodic phosphate was prepared, and the phosphoric acid estimated by the ammonium molybdate method. 25 c.c.

gave

(1) 0844 Mg2P,O, = '05398 P2O

66

*0540

5

(2) 0845 Estimations of phosphoric acid were made in varied quantities of the solution of sodium phosphate with the addition of potassium silicate, by the molybdic method.

The amounts of the solutions employed and the results obtained are given below.

0492 SiO2+1c.c. sodic phosphate =0022 P2O, gave 0023 P2O,

5

2

A solution of tricalcic phosphate in nitric acid containing in 50 c.c. 0379 grs. P,O, gave with 3100 SiO, 0381 grs. P2O,

⚫3100

1-0000 alum gave '0382 grs. P2O ̧

The ammonium molybdate and magnesium chloride solutions used in these determinations were made as recommended

by Abesser, Jani and Märcker in their paper on the estimation of phosphoric acid (Fresenius Zeitschrift, 12th year, p. 252), and all the operations were conducted as there advised. The above results show that in no ordinary case is a previous sepa ration of silicic acid necessary to ensure all desired accuracy in the estimation of phosphoric acid by the molybdic method. Prof. Kolbe's Laboratory, Leipzig, Dec. 17, 1875.

ART. XXVIII.—On the youngest Huronian Rocks south of Lake Superior and the age of the Copper-bearing Series; by T. B. BROOKS.

IN the summer of 1874, Chas. E. Wright and myself, while exploring the country west and south of the Menominee River about ninety miles from its mouth, under the auspices of the Wisconsin Geological Survey, observed a large granitic area, the north edge of which was bounded by dark-colored hornblendic and nicaceous schists of Huronian age, which I have since concluded are the equivalents of the youngest member of that series yet observed in the Marquette Iron Region.* The prevailing form was a medium to coarse-grained gray granite, with rectangular crystalline facets of feldspar. In places it passed through gneissoid granite to a true gneiss, which was once hornblendic, the schistose structure of which always conformed with the underlying schists.

The lithological character of this wide granitic belt bore so much general resemblance to the Laurentian rocks, which are extensively developed on the waters of the Sturgeon River in Michigan, 10 to 20 miles to the northeast, that we were disposed at the time to believe that some phenomena of folding or faulting had brought rocks belonging to that system to the surface in an unexpected quarter. Professor Pumpelly and myself, several years previously had observed, farther to the north and west, similar granitic rocks crossing the Michigamme and Paint Rivers (branches of the Menominee), presenting similar puzzling relations with beds known to be Huronian. This formation is noticed in my Michigan Report, vol. i, p. 175, and the probability of its being Huronian, and younger as well as lithologically different from any rocks then known to be of that period, is pointed out.‡

*The staurolitic mica schist, Bed XIX. of my scheme. See vol. i, pp. 83 and 130, Michigan Geological Report, 1873.

A few small granite dykes were observed penetrating the hornblende schists along the granite border.

It is not improbable that some of the granitic rocks S. W. of Michigamme Lake in the Marquette Region, may belong to the same horizon.