and Jobin), that of barium is 40 (Clarke), that of strontium 2-5418 (Matthiessen), and that of lead is 11:37 (Reich.) Hence the atomic volumes are respectively 25·8, 342, 34-3, and 18-2. These numbers approximate very closely to multiples of the lowest value for oxygen, 26. The number for lead requires no change whatever, that for calcium but 0-2, and those of barium and strontium but from 0-4-0-5, to produce the exact multiples. Whether this relation is accidental or not, it is impossible for me to say, but the coincidence is certainly remarkable.

Another similar relation we find between the atomic volume of the diamond, and those of certain metals which are often classed as tetratomic. The sp. gr. of glucinum is 2·1 (Debray), that of aluminum 2·56-2·65 (as given in the "Handwörterbuch), and that of thorium 7-657-7-795, (Chydenius). Hence their atomic volumes are respectively 6-7, 10-1-10-6, and 30-430.9. The atomic volume of the diamond is 34, that of glucinum wants but 01 of twice that, three times 3-4 is within the limits between which aluminum varies, and nine times 3-4, or 306, is within the limits for thorium. Cerium, according to Wöhler, has the sp. gr. 55, corresponding to an atomic volume of 167, which is 0-3 less than an exact multiple of that of the diamond. Perhaps this metal might be classed with the three above mentioned.

The remaining metals, as far as we have data, seem to find places in none of the previous series. Thallium, according to Crookes, has a sp. gr. of 11.81-11.91. Its atomic volume, then, is 17-2. Winkler gives the sp. gr. of indium as 7-421, which corresponds to an atomic volume of 10-2. For niobium, tantalum, lanthanum, didymium, yttrium, and erbium, we have no data whatever.

The sp. gr. of iodine is 4.947 (Handwörterbuch). This gives as its atomic volume the number 25.6, almost exactly two thirds of its value in its liquid compounds. The atomic volumes of bromine and fluorine in their solid compounds I have not been able to calculate in any satisfactory manner from the data at my command, since the values obtained did not agree with each other. The atomic volume of chlorine in solids, has been calculated by Kopp, who obtained two values; but between these and its value in the liquid state I have found no relations.

For hydrogen, the data at hand are quite complicated, not being in any connected series of compounds, and as far as I have been able to decide, it appears to have several values. However, in some of its solid compounds, hydrogen seems to

30

27

possess the same atomic volume as in the liquid state, viz., 5.5. If we calculate the specific gravities of cerotene, C2, H,,, and melene, C, H., upon the basis of this value for hydrogen, giving carbon the amended atomic volume of graphite, we shall get for the first the sp. gr. 0.846, and for the second 0.848. The specific gravities actually found for these bodies, as given in Weltzien's "Systematische Zusammenstellung," are respectively 0-861, and 0.890, a variation in the first case, of only 0.015, and in the second case of 0.042, from the calculated values.

Having thus compared the atomic volumes of many elements with each other, a question which now naturally arises, is, how far are the relations which have been found, significant, and how far are they accidental.

In work of this kind, great caution is necessary, for a slight prejudice in favor of the kinship of two elements might often lead one greatly astray. Thus, gold, silver, aluminum, and indium, possess almost exactly the same atomic volume, about 102, and yet no direct chemical relations are known to connect them, and they seem to belong to four different metallic groups.

Probably of a similar nature is the seeming relation between zinc and the platinum metals. Yet in spite of these dangers of error, some of the relations which are found between the atomic volumes of certain elements, cannot but be considered as deeply significant. The series of relations which I regard as fully made out, are the following: The multiple series of the alkaline metals; the series formed by oxygen, sulphur, selenium, and tellurium; the nitrogen series, and also that formed by carbon, silicon, tin, and titanium. The equal values for the platinum metals, and the similar case of the iron group. Of the series less thoroughly made out, or less easy to account for, I am inclined to regard that which connects the iron group with zinc and the cadmium group and possibly also with molybdenum and tungsten, as genuine. Also that series connecting the metals of the alkaline earth with oxygen. The series formed by the diamond, with glucinum and the kindred metals, and also the seeming relation between lead and oxygen, appear to me more questionable.

As a rule it will be seen that whenever an element possesses more than one atomic volume in its compounds, these different values are distinctly related to each other, and also to the value of that element in the free state. But between the different allotropic forms of an element, we find no distinct relations at all! If this be found to hold true, it will be plain that the different properties possessed by the same element in different

compounds cannot be accounted for upon the supposition that in one compound the element exists in one allotropic form, and in another compound in another.

One of the most important points to which I have been led, seems to me to be the very direct and simple relations connecting the atomic volumes of an element in two different states of aggregation. In the oxygen series, the most common value for each element in the solid state is almost precisely two thirds of that which it has in the liquid condition. Phosphorus, arsenic, vanadium, antimony, and bismuth seem to have in the solid state, values just half those which they possess as liquids, and carbon seems to follow the same rule.

The exceptions found under zinc and mercury, however, must be borne in mind, showing as they do, that the law is not a general one. But how far these relations between solids and liquids holds true, seems to me a matter of great interest, as possibly affording a clue to some general law connecting the different states of aggregation.

One more point and I am done. In many cases, even when the relations between the elements have been plainest, it has been necessary to slightly alter the atomic volumes actually found, in order to bring them in accordance with theory. Some of these alterations have been wholly within the limits of experimental error, yet others are in this respect more doubtful. Now it has often been suggested that the atomic volumes of solids, like those of liquids, should be compared at similar temperatures, and under corresponding circumstances. For solids, the similar temperatures are supposed to be the melting points. Now in many of those cases in which I have altered the values actually found for elements, it would be wholly unjustifiable for me to ascribe the variations from theory to experimental error. And yet the multiple relations are often extremely plain, and by their occurrence in so many different series, they in a certain sense confirm each other. Therefore, taking all things into consideration, I am firmly convinced that whenever the atomic volumes of the elements are compared under strictly similar circumstances, then the relations between them will be found exact. In other words, the variations from theory are probably due to the fact that the specific gravities from which we calcu late, were determined under dissimilar conditions. Concerning the cause of these relations between different elements, I shall hazard no conjectures. Not only are our data too incomplete, and unfit to form the basis of any theory, but it seems to me that until we know more of the nature of the elements themselves, it will be too early to attempt to frame any hypothesis whatever, concerning the cause of relations between them.

ART. XXVII.-Contributions from the Laboratory of the Lawrence Scientific School. No. 7.—On some Minerals from Newlin Township, Chester Co., Penn., described by Dr. Isaac Lea; by S. P. SHARPLES, S.B., Assistant in Chemistry.

A YEAR or two ago, Wm. W. Jefferis, of West Chester, Penn., gave me several minerals for examination which had been described by Dr. Lea in a paper read before the Academy of Natural Sciences of Philadelphia, April 9th, 1867.

The first of these that I examined was the Diaspore which Prof. Brush described in this Journal and of which an analysis is given on page 169, of Dana's Mineralogy. The diaspore was embedded in a mineral which Prof. Dana calls margarite (loc. cit.), and which Dr. Lea in his paper spoke of as emerylite. I found it to contain silica, alumina, potassa and water, with traces of iron, magnesia and soda in the following proportions:

The oxygen ratio as will be seen, is for R, R, Si, H, 1:8:5:11: 2, which corresponds with that given for damourite.

The mineral occurs in mica-like crystals, some of which are from one to two inches across. Some of the specimens in Dr. Lea's collection have well crystallized edges. In color it varies from white to greenish or yellowish. Before the blowpipe it fuses with extreme difficulty to a white enamel.

Lesleyite. This mineral is described by Dr. Lea as follows: "Fibrous or lamellar, sometimes inclining to massive. Color whitish passing into reddish. Hardness about three. Streak white. Before the blowpipe it parts with its water and becomes opaque white. Does not fuse with borax. Does not dissolve in muriatic acid. Under the microscope it presents no observable characteristics.

"Its gravity is greater than that of quartz. There is a disposition in the crystalline fibrous structure to diverge from a central point, to be stellate, and in one crystal before me, the radiating fibers are nearly four inches long."

Analysis of the white variety gave me silica, alumina, potassa and water with traces of iron, soda, lithia and magnesia.

Oxygen ratio for R, H, Si, H, 1:20:14: 3.

The reddish gave C. W. Ropper of Bethlehem, Penn.:

Oxygen ratios for R, H, Si, H, 1:97: 148: 3.5.

These minerals occur as a coating on corundum and seem to pass insensibly into each other. Prof. Cooke has quite good specimens of the brown variety that came from Sparta, N. York, where it also occurs coating the corundum.

Both varieties are evidently the products of the action of water containing alkaline silicates upon the corundum, and are more nearly related to pinite than to any other mineral species. For pinite, Dana gives the formula (R, R), Si, H, 3:4:1. For the white variety the ratio is 7:5:1, while the reddish gives 3:4:1.

Pattersonite. Of this mineral Dr. Lea says: "Basal cleavage imperfect, rarely if ever presenting a hexagonal prism, but disposed to present triangular plates which joining make a sub-tetrahedral mass. The laminae are not flexible and but slightly translucent. The color is metallic bluish gray, resembling zinc. The streak is grayish. Before the blowpipe parts with its water but does not exfoliate nor intumesce. With borax gives a black bead." I found it to contain silica, iron, alumina, magnesia, potassa and water, with traces of soda and lithia. The potassa was determined by difference owing to the very small amount of the mineral at my command.

Ratio for R, R, Si. H, 1:57: 66: 43; for (RR), Si, H, 3:3:2.

* Sharples.