The stoneware bed, which is layer 7 of the above section, increases to 3 feet in Mr. Cutter's eastern bank. It is a sandy, yellowish clay with many small mica scales, and differs from the top-sandy only in the amount of sand which it contains. When thrown into water it slaked moderately fast, and an air-dried sample required 28 per cent. of water for mixing. The air shrinkage was 7 per cent. and the average tensile strength 60 pounds per square inch. Its behavior in burning was as follows:

Burning test of stoneware clay, W. H. Cutter, Woodbridge.

The bricklets burned steel-hard at cone 5, and became viscous at cone 33. It, therefore, equals a good No. 2 fire clay in its refractoriness, but burns denser at all the cones at which both were tested.

South Amboy. The stoneware clays dug in the vicinity of South Amboy differ from the other clays of equal refractoriness of the same district in being more plastic, of higher tensile strength and burning denser at cone 8 or 10. As shown in Chapter VIII they belong to an entirely different bed geologically.

The No. I stoneware clay from H. C. Perrine & Son's bank did not vitrify until cone 30, and probably would not have fused lower than cone 32.

The two grades of stoneware clay sometimes occur alone or may be found together in the same bank and have to be sorted by hand picking, as at locality 81, or they may be interbedded with other clays, as at locality 77, where the following section was observed in September, 1901:

Section at H. C. Perrine & Son's bank, South Amboy.

I. Yellow sand, thin laminæ of clay,

4. Red-spotted clay. For foundry work,

2. Black sandy clay, similar to fireproofing clay, but more sandy,.. 3. Gray-white clay with decomposed pyrite specks (No. I stoneware clay),

6-15 feet

12 feet

3-6 feet

5 feet

6. Terra-cotta clay. Burns buff. More plastic than layer 5. Viscous at cone 26-27,

5. No. 2 stoneware clay. May develop a yellow-green stain on weathering, but this disappears on firing,

3-4 feet

4-6 feet

7. Sand cemented by iron oxide,

6 feet

The second grade of stoneware clay from this district is included in the next class.

SEMIREFRACTORY CLAYS.

This class includes the lowest grades of fire clay found in the Middlesex district as well as many other clays of similar refractoriness, but known to the trade by other names, such as firemortar clay, No. 2 stoneware clay, pipe clay, pressed-brick clay, wad clay, terra-cotta clay, etc., as well as some purely local terms, such as "blue top clay," "yellow top clay," etc., which have no significance except to the clay miner. They fuse between cones 27 and 30 inclusive.

Fire Clays.

Most of these low-grade fire clays become viscous at cone 27. They are probably more widely distributed both geographically and geologically than either of the two other grades which have been described from Middlesex county. On analysis they often show a higher percentage of fluxes as well as greater sandiness. Their air shrinkage is somewhat lower than the No. I fire clays, and their tensile strength higher.

Woodbridge.-One of these sandy clays is to be seen at locality 6 (Lab. No. 377), Anness & Potter's pit, and represents layer 4 of the following section, made September, 1901:

Section at Anness & Potter's bank, Woodbridge.

It is a whitish, gritty clay, with tiny mica scales, and slakes rather slowly. It is not highly plastic, and feels flaky and gritty when mixed with water. The air shrinkage was 4.6 per cent., and its tensile strength averaged 78 pounds per square inch. It behaved as follows in burning:

Burning tests of the top-sandy clay, Anness & Potter, Woodbridge.

The bricklets became steel-hard at cone 4, but the clay is only semirefractory, for it fused easily at cone 27. Its composition was as follows:

Chemical analysis of the top-sandy clay, Anness & Potter, Woodbridge.

This analysis is an interesting one when compared with that from Valentine's pit, locality 14, on p. 440. The difference in refractoriness is so large that it seems doubtful whether it can be due entirely to increased fluxes. It is probably caused in part by the higher silica contents of the clay.

Burt Creek. A so-called No. I blue fire clay belonging to the South Amboy fire-clay bed is dug in J. R. Crossman's bank, near Burt Creek (Loc. 65). Its position in the bank can be seen from the following section, made in September, 1901, near the western end of the excavation :

A test was made of layer 6 (Lab. No. 409), which is regarded as a good fire clay. It is lean, very fine-grained, and had a low tensile strength, viz., 20 pounds per square inch. Its air shrinkage was 6 per cent. Its behavior under fire was as follows:

Burning tests of a No. 1 blue fire clay, J. R. Crossman, Burt Creek.

This material is used as an ingredient of fire brick and stoneware mixtures. It becomes steel-hard at cone 5, but shows tiny iron specks. The sample fused at cone 29, so that its refractoriness is not great enough to call it a No. I fire clay according to the classification adopted in this report. As indicated on page 451, however, another sample of No. 1 blue fire clay from an adjoining bank (Loc. 66) did not become viscous until cone 33, so that there is evidently considerable variation in the fusibility of the No. I fire clays in this vicinity.

Chemical composition of a No. 1 blue fire clay, J. R. Crossman, Burt Creek.

This analysis is interesting as the ratio of silica to alumina is so low. It also shows that there are present nearly 4 per cent. of fluxes in the form of iron oxide and lime, and there is probably some titanium oxide present also, which, however, was not separated from the silica and alumina. Were it not for these impurities the clay would probably be highly refractory. It is interesting to compare this clay with the blue fire clay from the adjoining bank (p. 451), since the two resemble each other closely in appearance, yet differ so in fusibility.

Fire-mortar Clay.

Fire-mortar clay is a sandy type of clay found in not a few pits, and, as its name indicates, it is used for making mortar for setting fire bricks. A sample (Lab. No. 383) corresponding to layer 9 of the section on p. 436 was taken from Henry Maurer & Son's pit (Loc. 24, Pl. VIII, Fig. 2). It was a sandy, speckled, or finely-mottled clay with small mica scales, and was part of a bed about 3 feet thick. When tested it slaked fast and worked up with 28 per cent. of water to a mass, having an air shrinkage of 5 per cent. Its tensile strength averaged 99 pounds per square inch, and in burning it behaved as follows:

Burning test of a fire-mortar clay, Maurer & Son, Woodbridge.

It burned steel-hard at cone 5, and became viscous at cone 27. The chemical composition was:

Chemical analysis of Maurer & Son's fire-mortar clay.