The transitional elements titanium and vanadium on the one side of this series and copper on the other, are the outliers.

The second colored series contains the well known group rhodium, ruthenium and palladium. The third colored group contains the metals of the rare earths followed by the transitionals tantalum and tungsten. Finally the colored group of the platinum metals and gold. These and the remaining colored metals will be described in the next section.

One metal, zirconium, has proved rebellious to this classification.

The others have taken their places so easily and exactly that it seems as if there must be something inexact or incomplete. in our data respecting this metal. The most probable supposition seems to be the following. Zirconium nas but one degree of oxidation while the very closely allied metal titanium, has ions that are colored and colorless at different valencies.

Should zirconium prove to have a second degree of oxidation corresponding to colored ions, it would be brought into complete analogy with its congener and would find a place open for it in the tables.

In Table II it would take the vacant place immediately following titanium and between that metal and cerium. As a transitional metal it would take its place in Table III immediately before niobium in the second series. It is hardly necessary to remark that these are exactly the places for which its properties fit it.

All the elements contained in Table III have ions that function as kathions only.

A PERIODIC LAW OF COLOR.

It was necessary first to consider the elements in the great divisions into which they fall by reason of the color of their ions.

It now remains to consider the whole range of elements in one continued series from hydrogen to uranium.

Commencing with hydrogen (see Plate No. IV*) we have a double series of 18 elements with colorless ions only. Approaching one of the great colored groups which may be called the iron group we find two intermediate elements, titanium and vanadium which have both colored and colorless ions.

By their colorless ions they are united to the series which immediately precedes them in the order of numbers and by their colored ions they are united with the iron group which immediately follows. This iron group commences with the element chromium which in the numerical series immediately follows vanadium, so that after the transitionals titanium and vanadium each of which has at least one colorless ion, comes the group consisting of chromium, manganese, iron, cobalt and nickel metals which have colored ions only.

Approaching the next colorless series we find interposed the transitional element copper, a metal having the colorless cuprous and the blue cupric ions.

From this we pass to a colorless series commencing with zine and continuing with gallium, germanium, arsenic, selenium bromine, rubidium, strontium and concluding with yttrium. The ions of none of these elements show any tendency to

color.

Continuing in numerical order the next colored group consist of the metals ruthenium, rhodium and palladium. But in approaching these we find precisely as in the previous case two transitionals, molybdenum and niobium.

These are connected with the previous colorless group by their colorless ions and with the colored group next following by their colored ions. This colored group (Ru, Rh, Pd,) has colored ions only.

Continuing in numerica! order we approach the next colorless group. But as we pass from the colored to the colorless we find as before, a transitional, in this case, silver, which is connected with the previous colored group by its colored ions corresponding with Ag,O and Ag,O, and to the following colorless group by its ion corresponding to Ag,O.

In the plate the third and fourth colored groups should have been on the same horizontal line as the first and second.

The first of these colored ions is seen in the deeply colored hemi-salts of silver. Another may exist in the peroxide which dissolves in sulphuric acid with a dark green color.

From this we pass to the next colorless group of nine elements commencing with cadmium and ending with lanthanum. Approaching the next colored group we as before find a transitional element, in this case but one. At least but one is now known, but as we have now come to the region of little known metals of the rare earths it is possible that some one of those not yet thoroughly known may take its place alongside of cerium and thus bring this approach into complete symmetry with all the others.

Cerium connects itself with the colorless group immediately preceding by having colorless ions and with the colored group immediately following by its colored ions.

The colored group thus reached, composed of metals having colored ions only, consists of didymium, samarium and erbium. Then follow the transitionals tantalum (?) and tungsten.

Next, a series having ions colored at all valencies, namely osmium, iridium, platinum and gold. With gold the regular series terminates.

There follows what may be called the most curious part of the entire range of elements. This is found in the little group of six at the extreme end. In the principal series the colored groups are always immediately preceded and introduced by transitional elements, that is elements having both colorless and colored ions. The usual number of these transitionals is two. In the small final group the first two colored elements act as transitionals to the third. The first of the colored metals is thallium, this metal is allied to the alkalies by its thallious salts which are colorless; it is also closely related to the heavy metals, lead and mercury which are on each side of it. Even thallic sulphate and nitrate are colorless salts decomposed by water. But the thallic haloids form colored crystals and colored solutions and thus correspond perfectly to colored ions. Therefore thallium whilst chiefly related to the colorless elements on each side of it has nevertheless made a wellmarked step towards color by its single pair of colored ions.

The next colored metal, bismuth, has advanced much further towards color, for of its four valencies all but one have colored ions. It still retains its relation however with the colorless elements on each side of it, lead and thorium, by its one pair of colorless ions corresponding to bismuth trioxide.

Finally we have the last of all the metals, uranium, with colored ions at all valencies. Standing alone it occupies as it were the position of a group to which its transitionals, thallium and bismuth, lead up, and with it the series of the elements

closes.

Amongst the conclusions to be drawn from the facts that have been mentioned is this, that the color of the elementary atoms is to a large extent a function of their atomic weights. We find that with atomic weights,

From 1 to
From 52 to
From 65 to
From 103 to
From 112 to
From 145 to
From 192 to

47 the atoms are always colorless
59 they are always colored
90 they are always colorless
106 they are always colored
139 they are always colorless
169 they are always colored
196 they are always colored.

Elements whose place in the numerical series falls between these periods, have both colored and colorless atoms.

The six metals that remain are as we have seen, alternately colored and colorless.

Ostwald remarks in his great Lehrbuch that when the properties of the elements shall show themselves to be functions of their atomic weights, we have next to seek in the latter the cause of the former, and then we shall hardly be able to avoid the conception of a single primordial form of matter as suggested by Crookes, a form whose varied modes of agglomeration condition the various kinds of matter (Vol. I, p. 138).

Perhaps the facts in this paper described may be found to make a step towards this great end.

With the aid of the Arrhenius theory it has been possible to establish the principle that the colors of the atoms are those which they show in dilute solutions of electrolytes, and that the colors of elements are comparatively of little importance. In the second part of this paper there will be given incidentally a proof of the correctness of the dissociation theory from a new direction. In that part will be considered the combinations of atoms and two laws controlling in certain cases the interaction of ions.

ART. XXIX. Further Notes on the Gold Ores of California; by H. W. TURNER.

SOME brief notes were published in this Journal on the gold ores of California in June, 1894, and the following may be considered as an appendix to that article.

Gold in barite.-During the past summer, the writer examined some gold veins on Big Bend Mountain in Butte County, California, and found that one of them was of an unusual

character. The vein is known as the Pinkstown ledge. It is located about a half mile due south of the highest point of Big Bend Mountain (Bidwell Bar atlas sheet). The ledge strikes N. 13° W. and dips at a high angle (about 80°). It is from two to three feet wide where best exposed at the north end, and is composed of a soft heavy mineral, some of which is coarsely crystalline, with a granular structure, but most of it is finer grained with a schistose arrangement of the granules. No single crystals of the mineral were noted having a greater maximum diameter than five-eighths of an inch. Some of them show plainly a characteristic cleavage. Dr. Hillebrand made a chemical examination of this soft mineral and reported it to be barite. Three sections of the barite were examined microscopically, and these show that when fresh there is scarcely any impurity in the mineral, and in fact no other substance was noted except scattered minute reddish opaque grains which as seen under the microscope are reddish-yellow by reflected light, without metallic luster. They may be limonite. Many of the barite grains show distinct cleavages which appear in the thin sections to intersect at nearly right angles. A tendency to a radial structure like that of epidote was noted at several points. The relief of the barite is rather high. A sample was examined for gold by Dr. Stokes, who reported that "the barite contains gold but too small in amount to be determined in the wet way." There is said, however, to be enough gold in the deposit to pay to work, and the writer understood that the owner of the ledge obtained gold from it by grinding up the ore in a hand mortar, and panning it.

A considerable part of Big Bend Mountain, as exposed along the road from the bridge over the west branch of the north fork of the Feather river to the abandoned village of Big Bend, is made up of clay slates probably Paleozoic in age, with layers of greenstone schists, representing original augitic tuffs. The rocks along the east and south base of the mountain as seen along the river (the north fork of the Feather) are almost entirely greenstones, with one or two layers of sedimentary mica-schists. These greenstones are largely amphibolitic rocks representing original surface lavas and tuffs, probably augitic porphyrites, but now containing little or no augite. exact nature of the schist enclosing the barite vein was not determined. The south extension of the Pinkstown ledge owned by Clarke was examined but no barite was found, the rock on the dump being a white, fine grained schist, with a greasy feel. This as seen in this section is composed chiefly of minute, brightly polarizing fibers, perhaps tale, with numerous minute cubes of pyrite, arranged in rows.

The