of feldspar A and 25 per cent quartz. A comparison of this data with the calculated norm composition shows that the discrepancy is not very great.

Graphic granite B was of medium size grain. A block of this material was crushed, cleaned, and pulverized in the same manner as graphic granite A. Specimens of the pure feldspar and the graphic granite and a series of feldspar-and-quartz cones were prepared and tested.

Pure feldspar B deformed with cone 8. Graphic granite B deformed with the mixture consisting of 70 per cent feldspar and 30 per cent quartz and at the temperature of cone 9. A comparison of this data with the calculated norm composition indicates a discrepancy of 2.4 per cent, as the calculated content of free quartz is 27.6 per cent. Graphic granite C was coarse grained. A block of this material was prepared and tested in a manner similar in all respects to that followed with graphic granites A and B.

Pure feldspar C deformed midway between cones 7 and 8. -Graphic granite C deformed between cones 8 and 9, but nearer cone 9 temperature. It deformed with the mixture of feldspar C and 30 per cent quartz. A comparison of this data with the calculated norm composition shows the synthetic mixture to be 2.6 per cent high. This method of determining the quartz content of a pulverized feldspar is shown to be perfectly practical within general limits, but can not be depended on for determinations closer than 2 or 3 per cent. This discrepancy is possibly in part due to the fact that the quartz used in the synthetic mixtures was finer grained than the quartz contained in the ground graphic granite. This fact was proven by microscopic study of the fused cones, although the powdered materials were all thought before use to be of sufficiently uniform fineness.

In nature graphic granite is a mixture of feldspar and quartz in intimate crystalline relation, but both exist as distinct minerals. These graphic granites must be reckoned as mixtures of feldspar and quartz, which fortunately are remarkably constant for graphic granites containing feldspars free from lime and high in potash and low in soda. If the user of graphic granite will recalculate it in terms of feldspar and quartz and introduce it as such, much of the difficulty arising from its use will be overcome.

NORMATIVE COMPOSITION OF GRAPHIC GRANITES.

The compositions of 18 graphic granites, calculated into their mineral components as norms," is given in the table following. This list, with the exception of Nos. 3, 4, 8, and 14, is discussed by Bastin."

a For a discussion of methods of calculating the normative composition of a rock from the actual composition, as shown by chemical analysis, and the significance of the relationship, see Clarke, F. W., The data of geochemistry, U. S. Geol. Survey Bull. 491, 1911, pp. 404–406.

b Bastin, E. S., Geology of the pegmatites and associated rocks of Maine: U, S. Geol. Survey Bull. 445, 1911, pp. 41-42.

The table discloses the fact that the ratio of potash feldspar to soda feldspar in graphic granite is generally greater than 2:1 and often reaches 3:1. In those samples in which soda feldspar is the prevailing type of feldspar the potash feldspar is present in very small amount only. The graphic granites 15, 16, 17, and 18 belong to a class entirely distinct from the others. A fact worthy of note is the small variation in norm content of potash feldspar in Nos. 1 to 14, inclusive. With the exception of Nos. 2 and 13, the potash feldspar content is between 47.4 per cent and 55.2 per cent of the graphic granite, whereas the soda feldspar in the same series varies from 10.6 per cent to 29.2 per cent. In the first 14 samples-that is, the graphic granites high in potash feldspar only two contain more than 27.6 per cent free quartz and only two contain less than 23.2 per cent. Thus 70 per cent of the graphic granites studied contain 23.2 to 27.6 per cent of free quartz, a variation of 4.4 per cent.

PROSPECTING AND SAMPLING.

Too much emphasis can not be placed upon the importance of thoroughness in prospecting for feldspar and in sampling the deposit when located. Practically all the failures in feldspar quarrying are due to improper prospecting and sampling, which has led to an overestimation of the extent of the deposit and of the quality of material obtainable. In many instances the analysis of hand specimens taken in the course of ordinary prospecting have been made the basis of large investment and expenditures where the material of the grade sampled is so limited as to hardly justify a passing glance.

In prospecting for feldspar the geology and topography of the district must be carefully considered. If the country rock is more easily eroded than granite or pegmatite the dikes or sills are exposed above the surrounding country. If the country rock is more resistant to erosion than granite or pegmatite, the existence of the dikes or sills is indicated by depressions. Throughout the New England and North Appalachian States the former case prevails and in many places pegmatite rich in feldspar is exposed along the ridges or along the steep slope of the hills.

When such an exposure is found or a deposit of pegmatite is in any way discovered, the prospector should first ascertain the general nature of the walling rocks. From a hasty survey the existence of folding or faulting can be determined and also the deposit classified as dike or sill. An exposure may be either across the dike or along its strike and what may seem to be the face of a broad lens may be in reality only a thin sheet. With the formation identified as dike or sill the prospector should proceed to determine the strike and dip as well as the general dimensions of the deposit. To do this the

overburden should be removed from the exposure until the full width of the deposit is bare. With the data thus obtained the apparent strike should be projected 50 feet and the deposit uncovered at that point. Sometimes this projected continuation of the dike will prove erroneous, hence it is advisable where the overburden is heavy to locate the dike first by drilling. Prospecting should be continued at intervals of not more than 50 feet in both directions from the original exposure, and if the deposit seems to vary greatly in dimensions the intervals should be reduced to 25 feet.

The location of the dike by drilling is only of value as a labor saver in determining where the cuts shall be made. The reason for this is obvious. The drill sample can merely indicate that the formation is of a granitic nature and shows little as regards size of grain or the extent to which impurities exist.

After locating the limits of the deposit therefore open cuts or at least a pit should be made in order to expose a reasonable area of the surface. These exposures should not be farther apart than 50 feet. The cutting of a trench across the entire deposit as compared with sinking a pit through the overburden has been proven by experience to be little if any more expensive, as the trenching can be done with plow and scraper, whereas in digging a pit most of the work is necessarily hand labor.

Which surface is exposed by this method depends on the nature of the deposit, if it is a dike the surface exposed represents the crosssectional face of the deposit; if a sill the surface exposed is one of the walls and does not in any way represent an average of the deposit. The mode of procedure therefore depends upon the nature of the deposit.

SAMPLING A DIKE.

To sample a dike from the trenches cut, the most satisfactory method is as follows: Lay off the dike from wall to wall in 5-foot centers, beginning at a point 1 foot from the foot wall and placing the last center not more than 3 feet from the hanging wall. At each of these centers drill a row of three holes, 3 feet deep and 2 feet apart, on a 45° slope toward the hanging wall, the rows being at right angles to the trench wall. Charge each hole with one stick of “60 per cent" or one and one-half sticks of "40 per cent" dynamite, and tamp well with soft clay after attaching the detonator or exploder wire. Fire by whatever method is most convenient, but fire the rows in regular order, beginning with the one next the hanging wall. Much better results will be obtained if all the holes of a row are fired simultaneously. Each row should be carefully sampled and all loose débris cleaned away before the next round of shots is fired. As soon as the shots are fired, the loosened material should be removed and broken into pieces of approximately 3 inches diameter or less.

This material should include all that is loosened by the shot, but judgment should be used in handling it-for instance, a block of mica which was imbedded in the mass, but was entirely independent of the remaining material, should not be crushed and mixed with the other material loosened by the shots.

A careful record of the surfaces exposed should be made and especial emphasis should be placed on the gradation from the surface inward as regards size of grain and presence of associated minerals.

The material loosened by a row of shots should all be piled on a sampling cloth or wooden platform and thoroughly mixed, quartered, and two opposite quarters discarded. After mixing the remaining material thoroughly, it should be quartered, and two opposite quarters discarded. What remains on the cloth should be broken into pieces not to exceed 2 inches maximum diameter and the whole thoroughly mixed together. By repeated quartering this mass should be reduced to a sample of about 5 pounds.

In quartering especial care should be taken that the quarters discarded are entirely removed, because if some fine material is left on the sampling cloth each time, as is sometimes done by careless samplers, the resultant 5 pounds will contain an excessive amount of this fine material, which may have a different composition from the coarser material.

If the faces exposed by the shots indicate that the dike material improves rapidly with depth it may be advisable not to sample the rock next the surface, but to obtain the sample from the bottom of the holes. This can sometimes be done by striking with a heavy sledge the faces exposed in the bottom of the blast hole. If this is not effective a single hole may be bored in the bottom of the blast hole. This drill hole should slope 45° toward the foot wall, thus cutting under the face exposed by the first shot with the minimum depth of drill hole. A single hole 2 feet deep charged with one stick of "60 per cent" dynamite will generally move enough pegmatite to afford a good sample and this sample should be all removed and carefully sampled as explained above. After all the centers in all the trenches have been sampled in this manner, the depth of the dike is still to be determined. As this involves considerable expense for sinking shafts or cutting a tunnel, it may be advisable to first make or have made fire tests of the samples already taken. If the results of the tests prove satisfactory, the prospector is now ready to select a site for the test shafts or tunnel. The selection of this location will depend more or less on the topography of the country. If shafts are to be used at least two will be necessary and they should be at least 50 feet apart along the strike of the dike. One may be sunk at the natural place to open the dike, and the other should preferably be on higher ground. The shafts should be sunk not less than 25 feet into