the quality of the coal is regarded as dependent on the presence of neighboring igneous rocks.

Analyses of Matanuska River coals.

Proximate analysis.

"These analyses were made by F. M. Stanton on the same samples as those with corresponding numbers on pp. 60, 61. The differences in proximate analyses are due to the fact that the samples dried somewhat in the months which elapsed between analyses, and that somewhat different methods of analysis were used (see Bull. U. S. Geol. Survey No. 290, 1906, pp. 29-30).

Not determined.

The coals represented by analyses 2 to 17 are near the border line between semibituminous steam coals (as the name is generally used in the trade, but not according to Frazer's definition ") and bituminous coking coals. They are rather high in ash, but are otherwise of good quality. Analyses 5 and 6 bring the average ash abnormally high, but even aside from these most of the seams are so impure that mechanical separators or washing should be used to remove part of the impurities. The purer of these coals correspond very closely to the Crows Nest Pass (British Columbia) coal, which is the standard coking coal of western North America, and approximately to some of the coking coal of the East.

The bituminous coal of lower grade on Eska and Tsadaka creeks, which the analyses show to be intermediate between first-class bituminous coal and the lignite coals, should probably be called bituminous rather than lignite, its usual designation in the region. It is of about the same composition as the Franklyn, Black Diamond, and Renton (Washington) coal. It probably will not make good coke, but should serve well under stationary boilers, and possibly as a locomotive fuel and for other local uses. It should be especially well adapted to the generation of power by the use of producer gas in a gas engine.

Rept. Geol. Survey Pennsylvania, vol. MM, 1879, pp. 144-145.

Martin, G. C., Markets for Alaska coal: Bull, U. S, Geol. Survey No. 284, 1906, p. 29,

Carbon.

Nitrogen.

Oxygen.

Calories.

British ther

mal units.

COKING QUALITIES.

A rough test of the coking qualities of the coal from tunnel No. 2 on Chickaloon Creek (see p. 46) was made, during the summer of 1905, by coking a large pile of coal under a covering of stones and dirt. The resulting coke was hard and firm and had a good ring and a good texture. The test showed conclusively that a satisfactory grade of coke can be produced.

MINING CONDITIONS.

The dips

No serious difficulties prevent the mining of these coals. are so steep (10° to 60° on the anthracite, 18° to 85° in the east end of the bituminous area, and 20° to 44° in the west end of the bituminous area) that some method of stoping must be used. Miners who are accustomed to the steep dips in some of the coal mines of Washington and British Columbia will have no difficulty in this field. Drifts can be run from the level of some of the main streams, and enough coal found above drainage to supply the mines for some time. It will ultimately be necessary to resort to slope or shaft mining. These methods, or tunnels across the measures from the upper floor of Matanuska Valley, will probably have to be used on the anthracite very soon. But no operations should be undertaken in the anthracite area without thoroughly investigating the amount of workable coal, both by carefully exploring the surrounding country in search of outcrops and by testing the thickness and continuity of the seams by boring or other underground work. There is an abundant local supply of wood for building and mine timbering.

It will be necessary to wash or otherwise clean the coal from some of the seams. The percentage of ash can be reduced by washing from 10, 12, 15, 23, and 27 per cent to less than half and probably in some cases to a quarter of these figures. By washing tests at the coaltesting plant of the United States Geological Survey" the percentage of ash was reduced as follows:

Preliminary report on the operations of the coal-testing plant of the United States Geological Survey at the Louisiana Purchase Exposition, St. Louis, Mo., 1904: Bull, U, S. Geol. Survey No. 261, 1905, pp. 60-73,

GOLD.

DISTRIBUTION OF GOLD-BEARING ROCKS.

Gold-bearing rocks are found over considerable areas in the region. adjacent to Cook Inlet. A series of graywackes and slates, cut by small quartz stringers, occupies the eastern part of the Kenai Peninsula, forming the bed rock of the Sunrise mining district, extends across Turnagain Arm, and may be seen in the valley of Knik River still farther north. The search for placer gold in rocks of this type is warranted, and discoveries of commercial value may possibly be made.

North of Matanuska River, on the southern margin of a granite mass, occurs a band of highly crystalline mica schist. It is closely folded and infiltrated with fine quartz stringers. The only productive placer thus far developed within the region is situated on this belt of schists. It is noteworthy that the igneous rocks, granitic and volcanic, to which this series of mica schists gives place on the north and east, have so far proved barren of workable placers.

Farther northwest, near the mouth of Talkeetna River, a slateschist series, folded, intruded by granite, and containing abundant quartz stringers, again represents the gold-bearing series. The same rocks are also found just north of the Susitna-Cantwell divide. The bars of streams heading in this formation have for many years been worked with more or less success, and in 1898 were the scene of some local excitement. Several miles of claims were staked that year along the Chunilna, a tributary of the Talkeetna from the north. The goldbearing gravels of the newly discovered Yentna district are reported to have a slate bed rock, indicating the presence of metamorphic rocks to the west.

Though there is still question as to the relative ages of these series of rocks, there is little doubt but that their economic importance is due to the mineralization which has taken place during their infiltration by abundant quartz stringers. The concentration of gold values incident upon their wearing away by stream work has been the direct means of the formation of the placers.

The Pleistocene gravels at the head of Matanuska River are not likely to be gold bearing or of economic importance. Their glacial origin is distinctly unfavorable to the production of placer deposits. Their rapid mode of accumulation means a lack of sorting and concentration necessary for the formation of workable placers.

The present stream gravels are largely derived from reworking these older gravel sheets. Even with this partial reconcentration the gold content has not been increased sufficiently to pay for working. Productive placers appear to be limited to such streams as are directly

engaged in concentrating a metamorphic bed rock locally bearing gold, usually in the form of small mineralized quartz stringers.

Though placers have not been found within the areas of older volcanic rocks, mineralization has occurred. West of Hicks Creek a large cropping of gossan about 100 feet wide was found. This red capping is due to the oxidation of finely divided pyrite disseminated through a quartz porphyry. A sample selected for assay showed a trace of gold and no silver.

For several miles the whole southern flank of Sheep Mountain at the head of Matanuska River is colored a strong red from the oxidation of pyrite in the greenstones. At some points the sulphuric acid formed during the oxidation of the pyrite has bleached the greenstones to a pure white color. The rugged range thus tinted in vividly contrasting colors presents a marked scenic effect. Certain streams emerging from the range are so highly charged with iron salts as to color their gravels red with oxide. The pyritization of the greenstones, which are here roughly schistose, has affected a great thickness of rocks, but is of a diffused character. An assay of a sample selected as showing the maximum mineralization yielded only a trace of gold and no silver.

DESCRIPTION OF LOCALITIES.

WILLOW CREEK.

Placer gold is being mined in commercial quantities at only one locality within the area covered during the season, i. e., on Grubstake Gulch, a southern tributary of Willow Creek, which enters Susitna River about 30 miles above its mouth. Willow Creek proper was staked by M. J. Morris and L. Herndon is 1898, and it is reported that they extracted about $4,000. In 1899 A. Gilbert staked Grubstake Gulch, and in 1900 sold his interest to O. G. Herning, who manages the property for the Klondike Boston Mining Company, of Boston, Mass.

The valley of Willow Creek and the valleys of the small tributary gulches show clearly the results of ice action, the side streams occupying hanging valleys with steep gradients or falls where the smaller streams join the main water courses. Grubstake Gulch is an example

of such conditions. Near its mouth a rim of bed rock crosses the channel and is cut through by the present stream, which falls precipitously for about 150 feet in a very short distance and enters Willow Creek at a low gradient. An excellent dump for hydraulicking is thus afforded. Bed rock is a mica schist penetrated by small veinlets and augen of quartz. The schistosity at this point is N. 60° E., with a steep dip 40° to the north. The direction and inclination of the beds across the stream and the dip downstream are especially

13070-Bull. 327-07-5

favorable for the collection of any gold that might be concentrated from the rocks in the process of erosion which the valley has undergone.

In the last three years, 1904-1906, during which hydraulic methods have been in use, 900 feet of the creek has been worked out. Pay averages 200 feet in width with a depth from 2 to 3 feet. The gold is coarse and rough, and assays at the mint $16.58 per ounce. Very little black sand is found. The greater part of the gold occurs close to or in crevices of the bed rock, but it is not deemed necessary to clean up by hand, the hydraulic giant being relied on to sweep all gold into the boxes.

The wash, which is practically all confined to the gulch bed, there being no well-defined bench, is coarse, ill assorted, and not greatly waterworn. The many large bowlders make it necessary to employ at least two men in breaking up and removing oversize and add materially to the cost of extraction. Three Hendy giants are installed on the property, two No. 2 and one No. 1. Only one is used at a time, however. Seven hundred inches of water with a head of 180 feet is brought three-fourths of a mile down the gulch. A 24-inch pipe at the intake dam is reduced to 9 inches at the giant, to which is fitted a 3-inch nozzle.

The 900 feet of sluice boxes used are built entirely of whipsawed lumber. Bottom boards are 14 inches thick, side boards 1 inch thick, and the frames are 3-inch square timbers. The flume is 27 feet wide and 30 feet deep by inside measurements. Block riffles are used. A grade of 54 inches to 12 feet is maintained. The gravel is piped downstream into the boxes. Very little gold is caught below the fourth box, the greater part being retained in the second. Mercury is placed in the third, fourth, and fifth boxes.

The origin of the gold may be ascribed with certainty to the abundant quartz stringers in the mica schists. The coarseness and roughness of the grains suggest a near source of supply. It is very probable that the discovery of placer gold in commercial quantities in this region will be in the areas where mica schist is the dominant formation or where streams have eroded rocks of that type. The fact that placer gold has not been found in paying quantities where streams have headed in granitic or other crystalline rocks bears out this statement.

Recent prospecting has developed the fact that a well-defined bench about 75 feet above the bed of Willow Creek carries gold in commercial quantities. It is planned to install during the coming winter a hydraulic plant on the south side of Willow Creek, near the mouth of Wet Gulch, 2 miles below Grubstake Gulch, for their exploitation. The excellent facilities for dumping, and a catchment area at least as large as that of Willow Creek supplying water under suffi