red fuming nitric acid. After allowing them to rest until all gas evolution has ceased (which usually takes some six days), the solution is evaporated down at a gentle heat until the fumes of nitric acid are no longer perceptible. It is then very thick and syrupy. It is now diluted with water, and plumbíc carbonate is added in excess. The oxalate and undissolved carbonate are filtered off, and the solution slightly concentrated and allowed to crystallize. The glycerate of lead deposits in thick crystalline crusts. These are separated from the motherliquor, dissolved, and the lead precipitated out from the solution by sulphuretted hydrogen.

The colorless or light straw-colored filtrate is somewhat concentrated, and calcic carbonate is added to neutralization. The solution is filtered, if necessary, and to the filtrate is added 95 per cent alcohol. The calcium salts present are all precipitated, in greater part at once, and completely on standing twelve hours.

If the solution had been very concentrated the calcium salt is precipitated in a granular condition. If, on the other hand, it was more dilute, the salt only separates gradually, and has a beautiful micaceous and scaly appearance.

I had at first considered this precipitate to be pure calcium glycerate, but found on dissolving it in water, in order to free it from the lime and obtain the glyceric acid, that while the greater portion dissolved readily in warm water, a considerable portion, although not more than one-tenth of the whole amount, remained and dissolved only on continued boiling. This, when filtered off and washed in cold water, appeared as a dull white, almost impalpable powder, contrasting in appearance with the crystalline glycerate.

It was dried carefully at 100° until constant weight was obtained.

Calcium determinations were first made. Weighed portions were ignited in a platinum crucible once or twice with excess of concentrated sulphuric acid until the weight remained

constant.

5755 grms. salt yielded 4925 grms. CaSO, equal to 25-22 per cent Ca.

1759 grms. salt yielded 1505 grms. CaSO, equal to 25·16 per cent Ca.

The theoretical per cent of calcium in calcium tartronate is 25-32, while in calcium glycerate, allowing for two molecules of water of crystallization, it is 13.99.

I had analyzed the micaceous preparation of calcium gly. cerate about the same time, and had gotten in two determinations, 14.03, 14-07 per cent of calcium respectively. The difference was so great that I could not understand it. On

reckoning up the molecular weight, however, assuming one atom of calcium to be present, I got 159. The molecular weight of calcium tartronate is 158. Being dibasic, the molecular weight of the calcium compound is of course much less than the weight of the calcium compound of glyceric acid, a monobasic acid.

I endeavored twice to make a combustion of the salt in order to get the per cent of hydrogen and carbon. Each time calcium carbonate remained undecomposed at the heat of the combustion. I therefore gave them up.

I then took the remainder of my salt, grown rather small, to my great regret, and neutralizing the lime with oxalic acid, obtained the free acid. This, on concentration, deposited out crystals. On examination with a lens they were seen to be of tabular form, well agreeing with the appearance of tartronic acid obtained from nitro-tartaric acid. A combustion was made of these, and here, unfortunately, an accident to the potash bulbs lost me the carbon determination. The hydrogen determination, however, is given.

2

4348 grms. salt yielded 13-23 grms. H2O equal to 3.38 per cent hydrogen.

3

5

The theoretical per cent of hydrogen in C,H,0, is 3:33. An important test that I wished to make but was compelled to forego for the time, was to act upon this tartronic acid with hydrogen iodide. Were its structure symmetrical, it should yield a iodo-malonic acid, which by further treatment with HI or with reducing agents would yield malonic acid.

Wishing to obtain larger quantities of the tartronic acid for further examination, I have since oxidized another portion of glycerine and treated the products in the same way. This time I got no tartronic acid whatever, at least only a trace of calcium salt remained undissolved on heating with water. Evidently here the oxidation had proceeded somewhat differently as no tartronic acid formed. This result is not surprising on reflection, as the oxidation by nitric acid is not capable of much control, and a product once formed is liable to be still further oxidized. Thus glyceric and tartronic acids are both liable to be oxidized into oxalic acid, which always forms in considerable though varying quantity. Indeed the oxidation of glycerine by nitric acid is now known to yield a variety of products, of which, however, no doubt some are secondary

ones.

Thus Heintz has proved that racemic, formic, glycolic and glyoxalic acids are all found associated with the glyceric and oxalic acids in this product.

The tartronic acid just found, therefore, is only one of sev

* Ann. der Ch. und Ph., clii, p. 325.

eral smaller side-products. The known symmetry of structure of the molecules of all these side products, however, certainly argues in favor of a similar symmetry in the glyceric acid molecule.

There is one way of reconciling these two views of the structure of glyceric acid, and that is the assumption of the exist ence of two isomeric acids, of which one is normal and the other an unsymmetrical acid.

Some results that I have just obtained in purifying the calcium glycerate seem, indeed, to point this way. Should the unsymmetrical glyceric acid preponderate in this mixture, Wislicenus' reactions with hydrogen iodide are readily understood. Another fact, which should not be lost sight of, is that in the decomposition of iodo-propionic acid by moist silver oxide, Wislicenus* obtained not hydracrylic acid alone, but three other products accompanying it, so that the decomposition was not so simple

I am now engaged upon a study of this question, and hope to be able to give more information upon it in a short time.

ART. XIL-Note on the "Chloritic formation" on the western border of the New Haven Region; by JAMES D. Dana.

THE rocks of the hilly region west of the New Haven plain are, for nine miles westward, metamorphic slates, and beyond this distance mostly gneiss. Immediately adjoining the region there is what Percival has called a "chloritic formation," the area trending about north-northeast; then on the west of this, with the same trend, (2) a hydromica slate, but little removed from argillite, becoming slightly garnetiferous toward the western limit; next (3) a glossy garnetiferous mica slate, containing some beds of gray semi-crystalline limestone; next (4) at Derby, common gneiss and coarse porphyritic gneiss. These rocks are involved in one system of folds, and are throughout conformable in bedding.

The rock of the "chloritic formation" varies much in texture and composition in passing from the Sound northward. Near Savin Rock, on the Sound, it is a chloritic hydromica slate. The gray and slightly silvery surface is more or less blotched and lined with the olive-green of chlorite, and the rock has in the mass in general a greenish tint. The slaty structure is usually perfect, and yet some layers fail of it. Grains of magnetite are common, and, less so, those of pyrite.

* Ann. der Ch. und Ph., clxvii, p. 41.

This slaty variety of the rock continues with little change for a mile and a half north. Beyond, the massive layers increase in extent. At the deep Derby railroad cut, two miles north of Savin Rock, the massive variety constitutes more than half of the rock exposed in the sections; and it is not all in separate beds; for thick beds that are slaty in one part are in others for many rods massive, and it is impossible to separate the massive from the slaty by any stratigraphical planes.

This massive rock is commonly without a trace of bedding; at the same time, it is variously and extensively jointed, so that it affords only deceptive indications of strike or dip. It varies in color from greenish gray to dark olive-green and blackish gray. Some of it is almost cryptocrystalline; but in general the texture is fine granular. Part of it is porphyritic with small crystals of a whitish feldspar.

Between this Derby cut and "Maltby Park," a mile and three quarters west of north, this massive rock constitutes nearly all the outcrops; and in some places the porphyritic variety is pale greenish gray, from the thickly crowded feldspar crystals.

Over Maltby Park the rock is again slaty and silvery, often with blotches of chlorite-a chloritic hydromica slate-as at Savin Rock; yet with enough of both the ordinary and por phyritic massive kinds among the slaty layers to exhibit its close relation to the rocks farther south. The slate occasionally has the chlorite in large lenticular concretions, and now and then is light gray and contains crystals of pyroxene. In some places, especially along seams, it is epidotic. Veins and seams of quartz are numerous. In the slate there are interrupted beds of limestone. Part of the limestone contains serpentine and is a handsome verd-antique marble; and with the serpentine there are often grayish green cleavable pyroxene (sahlite), asbestus and chromic iron.

A mile farther north, or five miles from Savin Rock, (west of Westville), the rock is almost wholly a dark green chlorite slate the micaceous part absent.

The rocks in the course of the five to six miles are-in recapitulation-commencing at the Sound:

For 1 miles, chloritic hydromica slate, little of it massive. 1 to 2 miles, chloritic hydromica slate, much of it massive. 24 to 4 miles, massive chloritic rock with little of it slaty. 4 to 4 miles, chloritic hydromica slate, very little of the rock massive.

44 to 6 miles, mostly dark green chlorite slate.

It is to be noted that throughout the formation the slaty and massive portions are so associated, sometimes as alternating

beds, sometimes as parts of the same beds, that their common metamorphic origin cannot be questioned-a point that I have studied for years.

The resemblances of the massive rock to trap was long since noticed by Professor Silliman, who, in a paper on the geology of New Haven and its vicinity, published in 1811 in Bruce's Mineralogical Journal, called it "primitive greenstone." In fact, the similarity in external aspect is so close that hand specimens from some portions of it would without question be pronounced trap-that is, doleryte, diabase, or melaphyre-by the most experienced lithologists.

In 1872, an incomplete analysis of the feldspar in the pale grayish green porphvritic rock, outcropping just south of Maltby Park, was made, by Mr. Edward S. Dana. The amount of silica afforded by the feldspar having been found to be but 45 per cent, the conclusion suggested was that the rock consisted largely of labradorite, and that it was probably essentially identical with part of the trap of the Connecticut valley dikes. In view of the presence of chlorite, I hence regarded the compact rock of the region as a metamorphic diabase; and it is the rock specially referred to under that name in the last edition of my Manual of Geology.

Still, the analysis, besides being incomplete, was not satisfactory because the feldspar crystals, although of the normal hardness, were granular in texture, without good cleavage, suggesting that they might possibly have undergone a partial alteration. On account of Mr. Dana's departure for Europe, he was compelled to leave the investigation he had begun unfinished ; and so it has remained until this summer, when it was taken up, at my request, by the skillful analyst connected with the mineralogical department of the Sheffield Scientific School of Yale College, Mr. George W. Hawes. His results prove that the rocks are in fact metamorphic doleryte, metamorphic diabase, and metamorphic melaphyre; the first two, labradorite rocks, and the last an oligoclase variety. To distinguish these metamorphic rocks from the igneous of the same composition, they are named, on my suggestion, metadoleryte, metadiabuse, and metamelaphyre. The examples are part of a long series of rock species which have representatives both among igneous (or intrusive) and metamorphic rocks. Other kinds are dioryte and metadioryte. syenyte and metasyenyte, felsyte and metafelsyle, etc.

We have here the important geological fact that labradorite is a prominent constituent of certain metamorphic rocks which have the aspect of much dioryte, and which are probably of Lower Silurian origin.* The labradorite-a lime-and-soda feld

* On the question of their age I have collected many facts and propose before long to publish.