The observations made with weaker magnetic fields will be given with somewhat less detail :

In the above tables no corrections have been applied with reference to the facts, already discussed, that the test-coil is not used in just the same manner as the loops on the crosses, or to the further fact that the readings of the ballistic galvanometer scale are proportional to the tangent of twice the deflection instead of being proportional to the sine of one-half the deflection. In the table which follows, and which is in the main derived from those going before, corrections are applied with these points in view.

}

.928

4,900 (July 3,'88: no observations on tr'nsv'rse current.)1.02

AM. JOUR. SCI.-THIRD SERIES, VOL. XXXVI, No. 212.-AUGUST, 1888.

The last column shows that with fields of medium strength the relatively greater transverse effect in the thick cross corresponds closely to the more intense magnetic induction through this cross.

The difference between the second and third values for No. 5 in this column, and also that between the second and third values for No. 7, must be regarded as mainly accidental, a fact which shows the need of caution in drawing numerical conclusions from this series of results. It appears, however, that the transverse effect in the thin cross, No. 5, maintains a nearly constant ratio to the magnetic induction through the same when the magnetizing field is increased from about 9000 to more than 16,000; but that with a similar increase of magnetizing field the transverse effect in the thick cross, which attains an extremely high degree of magnetization, does not increase enough to maintain unchanged its ratio to the magnetic induction through the cross.

(To be continued.)

ART. XV.—Analyses of the Waters of some American Alkali Lakes; by THOMAS M. CHATARD.

IN the course of the geological examination, by the U. S. Geological Survey, of the Great Basin or country lying between the Rocky Mountains and the Sierra Nevada, samples of the waters of most of the lakes of that section were collected and sent for analysis to the laboratory of the Survey. Many of these analyses were published in Bulletin No. 9, U. S. Geol. Survey, but when I began a special study of the more important sources of natural alkalies, for the purpose of preparing the way for their practical utilization, a re-examination was, in some cases, deemed advisable and samples were also obtained from Owen's Lake, California, a locality not visited by the Great Basin party.

The four analyses here given represent the most important alkali lakes so far known and as the waters of two of them are now being utilized, it is believed that the information will prove of interest.

The analytical methods used are, in general, well known. The carbonic anhydride was, in all cases, determined by dis

tillation, being collected and weighed in a potash bulb. For the boric acid, in the water of Owen's Lake, the excellent method of Gooch* was followed; for the Mono and Ragtown waters the process of Stromeyer, with modifications of my own, was used, but, though giving very fair results, the process cannot be compared for ease, simplicity and accuracy with the former method.

The determination of the alkalies in such very dense waters was, at first, found to be difficult, owing to the large quantities necessarily taken.. To get determinations which shall be fairly accurate when referred to a liter, portions of not less than 50 c.c. should be taken, the water in all cases being carefully weighed. The amount of alkaline chlorides corresponding to 50 c.c. of such waters is very large, and, when the final evaporations are being made, there is great liability to loss, owing to the formation of a salt crust on the surface of the liquid and subsequent spirting caused by steam and ammonia vapor produced under the crust. This takes place no matter how carefully the evaporation be carried on, and many determinations were thus lost.

The use of hydrochloric acid enables us, however, to completely obviate this difficulty. The purified alkaline chlorides are to be evaporated till crystallization begins; the platinum vessel is then removed from the water bath and an equal bulk of very pure, highly concentrated hydrochloric acid added to the solution. A copious precipitation of finely granular salts at once ensues, and, on replacing the vessel on the water bath, evaporation goes on quietly and rapidly, no salt crust is formed, and when the final heating is given, little decrepitation occurs. In this manner, it is as easy to handle several grams of chlorides and to obtain accurate determinations as it is when we have to deal with the far smaller quantities usual in mineral analysis.

Of the four lakes to be considered, the most northern one is Abert Lake, in south-eastern Oregon. The sample analysed was "collected by H. T. Biddle at middle of west side of lake, one foot below surface, 30-40 feet from shore, September, 1887." The total quantity at my disposal was about 200 c.c., an amount too small for any extended research, but sufficient for all practical purposes. For each determination, 25 c.c. = 25-7792 grams were taken.

Specific Gravity, 1·03117 at 19.8°.

*Gooch, Proc. Am. Acad. Sci., 1886, p. 167.

Proceeding southwardly, we have next the "Big Soda Lake" near Ragtown, Churchill Co., Nevada. This, with the adjoining "Little Soda Lake," has been very fully described by King* and by Russell,+ and my own observations, during my short stay at this place, were confined to the technical aspects of the manufacture of carbonate of soda, which is carried on at both lakes.

The sample of the water of the "Big Lake," of which the analysis is here given, was collected by Mr. Russell in 1881, and was taken at the depth of one foot. The analysis was made in 1884 and was published in Bulletin No. 9, U. S. Geol. Survey, the CO, being then determined by difference. Since then it has been carefully determined in duplicate, with the following results.

Specific Gravity, 1.0995 at 19.8°.

Mono Lake, Mono Co., California, is next in succession. This locality is described at length by Mr. I. C. Russell in his paper, "The Quaternary History of Mono Valley, California,' which will shortly appear in the 8th Annual Report of the U. S. Geological Survey; for the present it suffices to say that this large body of water, of a composition so favorable to utilization, is, for practical purposes, inaccessible, and that the high altitude and consequent shortness of the evaporating sea

* King, Clarence, U. S. Geol. Expl. 40th Parallel, vol. i, pp. 510-513.

† Russell, I. C., Geological History of Lake Lahontan, U. S. Geological Survey. Monographs XI.

son would, under any circumstances, render the success of any industry established there very doubtful.

The sample analyzed was collected by Mr. Russell in 1882 and was taken at the depth of one foot, on east side of lake. Specific gravity, 1·045 at 15·5°.

Finally, at the southern end of the series, we have Owen's Lake, Inyo Co., California. This lake has been described by Oscar Loew in Wheeler's Report for 1876, pages 189–194, and upon it most of my field work for 1886 and 1887 has been done. Its greatest dimensions, as given by Loew, are seventeen miles long by nine miles wide, its greatest depth fifty-one feet, and it contains, according to his calculation, 22,000,000 tons of carbonate of soda. The Inyo Development Co. has begun the manufacture of soda at this point and to it I am indebted for every needed facility for making observations on the evaporation of the water and on the various salts obtained by its fractional crystallization. These products, together with others since obtained, are under examination and the results will be published later.

The sample analyzed was taken by myself, Sept. 17, 1886. Portions of 100 c.c. were taken for each determination, the duplicates agreeing very closely.

Specific gravity, 1·062 at 25°.

In 1 Liter. Per cent.

(CaMg CO 3

·055

Mg *005

Fe2O3 ·014

·02

NaCl

Na2SO4

Na, CO 3

NaHCO3