6. That when there are several such derivatives, the Greek numeral prefixes di, tri, tetra, etc., may be used to indicate the number of molecules of water removed from the ortho acid to yield the meta-form.

7. That intermediate between the simple ortho- and metaacids, are others containing more than a single atom of the negative radical; and that these acids may be designated by di, tri, tetra, etc., (according to the number of negative atoms) prefixed to the name of the acid, while the number of molecules of water removed from a multiple of the normal acid to form them is indicated by the same numerals prefixed to the meta.

8. That while the negative atoms in the compounds just mentioned are united by oxygen, there may be other compounds whose negative or positive atoms are united directly; thus producing a fourth class of acids and of bases.

By classifying thus the substances known as acids and bases, -and of course the salts derived from them-it is hoped that their relations to each other may be made clearer. And by giving them systematic names, their position in the series may be fixed, and a step be taken toward the establishment of a rational nomenclature.

New Haven, Oct. 10, 1867.

ART. XLIII.—Crystallogenic and Crystallographic Contributions; by JAMES D. DANA. On the Feldspar group of Minerals; supplementary to ART. XXIX, page 252: On the Chemical Formulas of the Silicates.

1. Chemical formulas of the Feldspars, and of some other Silicates.

THE resemblance or affiliation between the species of the Feldspar group is of the most intimate kind. There is not merely isomorphism (or close approximation in crystalline form), which, when alone, is of little value in determining relations, but also, apparently, the profoundest isotypism. For the approximation includes near identity in (a) the habitual forms; (b) the direction and facility of the two cleavages; (c) the hardness; (d) the color, and tendency to opalescence; (e) all optical characteristics; and as regards chemical composition, they agree in consisting of silica, alumina, and an alkali or alkaline earth, to the exclusion of magnesium and iron; and the oxygen ratio between the protoxyds and sesquioxyds is for

The expression "oxygen ratio between the protoxyds and sesquioxyds," is equivalent, under the new system, to the ratio between the combining powers of the protoxyd or dyad and sesquioxyd or pseudotriad metals," the so-called dyad

each 1: 3. The group, therefore, is one in nature, and chemistry must have some explanation of this unity. The following observations bear on the question of its origin.

1. The oxygen ratios for the protoxyds, sesquioxyds, and silica, in the different Feldspars, are as follows:

[The ratio 1 : 3 for the basic oxyds is expressed in the formula R+A1, or (R3+31); and, for the basic metals, in the formula R+33Al, or (R+38AI); or if the perissads are distinctly indicated, in the formula (R, R2)+3ẞÂl, or (†(R, R ̧)+ BAI).]

Besides the above-mentioned ratios, some analyses afford 1: 3:41, 1:3:5, 1:3:5, and other ratios between 1: 3: 4 and 1 3 6; and so also intermediate ratios between 1: 3: 6 and 1:38; between 1: 3: 8 and 1: 3:9; and between 1:3:9 and 1: 3:12. But these cases, (and also those in which there is some variation from the ratio 1: 3,) are usually regarded as exceptional to the above as the true normal ratios; and they have been attributed to the following causes:

(a) Incorrect analyses.

(b) Impurities and often mixtures of different feldspars through inter-crystallization, some remarkable examples of which have been detected by microscopic and optical examinations.

(c) Alteration; caused either (1) by the infiltration of ordinary waters carbonated or not-such waters filtered through powdered feldspar being known to become alkaline in a short time, and the rocks containing feldspars having been exposed to this action through long ages past; and examples being numerous of the alteration of feldspars, ending in their kaolinization, in which both alkalies and silica are lost; or (2) through the same process aided by mineral ingredients in the waters, resulting in the introduction of magnesia, oxyds of iron, etc., and in other changes.

As regards the views here presented, it matters not whether these intermediate ratios are irregularities produced by the above methods, or are normal ratios.

group here including not only Ca, Mg, Fe, etc., but also K2, Nag, Li2, etc., as illustrated in the classification on p. 204 of the former paper. I use beyond the expressions of the old system, as they are intelligible to all chemical readers, whether versed in the new system or not.

2. The oxygen ratio in Anorthite is that of a Unisilicate, for 13:41:1; and the formula may be, on the old and new system:

In the other species, the oxygen ratios vary from that of Anorthite, through an increase in the proportion of silica, they being as follows:

3. The silica of a Unisilicate is enough for the making of a feldspar; that is, to produce a species having that assemblage of crystalline, physical and other qualities which characterize the group. The type is therefore essentially Unisilicate. But while Unisilicate, it admits of increase in the proportion of silica to produce its variety of species.

4. An excess of silica above that of a Unisilicate may exist under the Unisilicate type in one of the two following conditions, and still be not an impurity.

A. It may be basic silica; in which case the general formula for the group would be

R(or R,R,) in Labradorite, Andesine and Oligoclase = Ca, Na, with occasionally some K; in Hyalophane, which has the formula of Andesine, Ba and Na; Ba replacing the Ca of other feldspars. These formulas are partially presented on page 260.

=

B. It may be accessory silica; by which is meant, not an ingredient unessential to the species, but a true component, accessory to the fundamental type of the series. The formulas on this view would be:

*This idea of the feldspars is recognized by the author in the last edition of his Mineralogy (1854) in which he heads the group as follows: "Type ratio for the Oxygen of bases and silica 1: 1, and varying from this by the addition of silica to the type"; and gives the following as the formulas:

According to either of these two methods, A or B, any other proportions of silica exceeding that of the Unisilicate may be expressed in the same manner as the above.

The last of the formulas under A, if expanded, becomes,(R3)2 Ši3 + 3Ä12Si* + 4Si Si®

Old
New

3

SiO(R2)2 + 3[Si}0,【BAI,] + 4[Si}0,[ySi2]

in which the last member consists of silica alone; and thus the formula approaches the corresponding one under B, yet with this important difference, that the silica in this member is in the former in two states, a basic and an acid.

5. The excess of silica above the amount required for the Unisilicate increases with the increase of the alkalies, or the perissad elements, in the base.

In Anorthite, true unisilicate feldspar, the base consists of lime without alkalies. But in the other feldspars, with higher silica, it is partly soda or potash, with lime, or in one case (Hyalophane) baryta; or it is wholly soda or potash. The ratios of non-alkaline to alkaline constituents in the bases in these species are, for the average results of analyses, as follows:

Thus the above statement is fully sustained by the average reresults of the analyses. Looking to special analyses, it will be found that in general those Labradorites which contain less than the normal proportion of silica (and so approach Anorthite) are those which contain most lime in proportion to the alkali; and those which give the oxygen ratio 1:3: 65, 1:3:7 are those which contain least lime relatively to the amount of alkali; and so for the ratios between 1:3:8 and 1:3: 9, and others between 1:3:9 and 1:3:12. There are exceptions to this statement, as should be expected when we consider the diversity of causes (see p. 309) which may have aided in producing these intermediate ratios. But the cases in accordance are sufficient to sustain it.

It would seem, therefore, that this excess of silica is dependent in part, if not wholly, on the alkaline or perissad nature of the base.

There is a prominent exception to this view in the species Leucite, which, although isometric in crystallization, is regarded as of the Feldspar group, since it has a similar general constitution, with the Andesine ratio 1:3: 8. For although it is thus related to Andesine, the protoxyd base is solely alkaline. Yet, as it stands apart in its isometric crystallization, it may not be as certainly in opposition as would at first seem. It may belong with Sodalite and Haiynite to an independent series in the group. Sodalite (with Haüynite) has the anorthite ratio 13: 4, and is a soda species; while Leucite, with 1:38, is a related potash species; and if the two are in an independent series, it would appear that potash is more strongly determinative of the higher ratio of silica than soda. 6. Facts exist among other silicates that throw some light on this subject.

A. The Mica group has the following for the oxygen ratios between the bases and silica:

Now the ratios of the non-alkaline to alkaline constituents of the protoxyd bases in these species, as deduced from the average results of analyses, are as follows:

The non-alkaline part of the protoxyd base in the species is mostly magnesia and protoxyd of iron, with some lime; and the alkaline, mostly potash, except in Lepidolite and Cryophyllite, which contain also lithia. The protoxyd bases are most exclusively alkaline in Cryophyllite.

The parallelism with the Feldspar group is striking; and there may be the same reason, therefore, for the increase of the silica, and not that mentioned on page 259.

B. Again, among Unisilicates, the Meionite section of the Scapolite group illustrates the same principle. The species Meionite, Mizzonite and Marialite are closely alike in crystallographic and physical qualities, so much so that Mizzonite has generally been made a variety of Meionite; and Marialite if it had been as long described would probably have shared the same fate. (It was first announced by vom Rath last year in the Zeitschr. d. deutschen geol. Ges., xviii, 635).