I note, incidentally, that the almost identical agreement of the values of \(H\), computed from V. D. Dudetskii’s Pavlovsk observations, both at low and at high temperatures, with the aid of the temperature coefficient given in the instrument certificate, with the values of \(H\) according to the observatory instruments, may be regarded as an independent determination of this temperature coefficient, giving the same results as the Washington measurements.

Four series of determinations of the induction coefficient, made by V. D. Dudetskii at Pavlovsk, gave for the quantity \(h\) the values 0.0104, 0.0104, 0.0109, and 0.0095, very close to the value 0.0107 indicated in the instrument certificate.

In conclusion I shall point out that a detailed analysis of all sources of errors leads to the conclusion that the error in the values of declination reduced on the average to 1914 is of the order of \(2'—4'\), depending on how accurately the latitude of the point is known; that the error in the values of \(H\) is 0.0001 C.G.S. for points where both oscillations and deflections were observed with Bauer’s instrument, 0.0002 for points where only observations of oscillations were made, 0.0002 for the determinations of Meshcheryakov and Butakov, and 0.0004 for the determinations of the remaining observers, whose instruments were less carefully compared with Bauer’s instrument; finally, that the error in the values of inclination does not exceed \(2'\).

B. P. Weinberg.

Distribution of Magnetic Declination in the Yenisei Governorate.

In view of the extreme insufficiency of magnetic determinations made and published in Siberia in general and in the Yenisei Governorate in particular, in 1914–15 I made an attempt to collect the results of determinations of magnetic declination made for practical purposes by various agencies and consisting, in the majority of cases, of indications of the magnitude of the declination on plans and maps of one or another tract. Although such indications are often not the results of actual measurements, but merely repetitions of similar indications on old plans of the same tract or on plans of neighboring tracts, and, even if based on actual determinations, are of little accuracy, nevertheless I resolved upon this attempt in view of the absence of any other data for many parts of the Yenisei Governorate (especially before my determinations of 1914 and 1915) and in the hope that a large number of even little reliable values of declination would compensate for their poor quality and even, in some cases, their dubiousness.

Thanks to the courtesy of representatives of various agencies, I succeeded in collecting all, apparently, materials of this kind, which came to me in the form of 365 values of magnetic declination, taken from the documents of the land-survey archive by the Yenisei provincial land surveyor P. S. Onykienko; 177 values of declination supplied by the head of the technical section of the Yenisei Resettlement Administration, Ya. Braikovskii; 36 values of declination according to data presented by district mining engineers; 7 values determined by parties investigating the Yenisei River; and 2 values reported by the Mansk forest ranger.

In view of the impossibility of separating those of these data which represent the results of direct determinations from those which are not such, and also owing to the impossibility of a comparative evaluation of their accuracy, I decided to make use of all these valu-

declination values without distinction. To these I added considerably more reliable material, to which I also assigned greater relative weight, namely the values of the declination determined in the region of the Yenisei Governorate up to 60 degrees north latitude from special magnetometric investigations: 8—Fritsche, 1867–1883; 1—Sharnhorst, 1873; 1—Miller, 1875; 15—Ossipov, 1902; 12—Weinberg and Dudetckii, 1912; 31—Weinberg, 1914–1915. I left the more northern part of the governorate’s territory without a compilation of declination values, in view of their extremely insufficient number.

Since the declination values taken into account cover the time from the end of the 1860s of the last century to the present (I also left aside the still earlier determinations of several travelers), it was necessary to reduce them to a single epoch; for this purpose I determined the secular variation of the declination for various parts of the territory investigated—also with great approximation, in view of the scantiness of the material that could be used.

I divided the declination values reduced to the epoch 1910 into groups corresponding to areas bounded by meridians at intervals of \(1^\circ\) and by parallels at intervals of \(40'\) (almost squares, each with an area of about 5 thousand square versts), and took weighted means from the values of each group. The values obtained, despite the shortcomings of the initial material, showed a clear decrease in the absolute magnitude of the declination (eastern, negative) as one moves eastward and southward. Taking, as a first approximation, a linear change in both directions and applying the method of least squares, I obtained the following formula expressing the declination at various points for 1910:

\[ \delta=-\left[8^\circ 51' + 15'\frac{\varphi-55^\circ 40'}{40'} - 18\frac{\lambda-93^\circ 00'}{50'}\right], \]

where the mean error of the first term was found to be \(\pm 3'.4\), that of the coefficient of the second term \(\pm 1'.5\), and that of the coefficient of the third \(\pm 1'.8\). The distribution of deviations from this formula does not reveal any regularities, with the exception of the southern belt of the governorate, where in the middle part the declinations are about 10 minutes greater than those obtained from the formula, and in the eastern part about 10 minutes smaller; but, owing to the insufficiency of the material, it is difficult before new measurements to assert that these deviations indicate an actual departure of the mean declination values from a linear distribution.

In any case, the present attempt has made it possible to expand considerably our knowledge of the distribution of declination over the extent of the Yenisei Governorate and, what is still more important, has allowed us to hope that if similar material could be collected for other governorates as well (which presents no small difficulties), then our extremely scanty information on terrestrial magnetism in Siberia could be supplemented by it—information so necessary for a number of practical tasks.

The labor of processing the data that served as material for the present preliminary communication was shared with me by S. I. Shubin.

17.10.18.

Addition (1. 6. 23). The assumptions expressed have been justified, and the 3,279 collected declination values, processed in the form of the article (1920; more than 6 printed sheets) “Compilation of magnetic determinations made in Siberia from 1820 to 1918. Part II. Unpublished determinations of declination made by various agencies,” served as good support in determining the normal distribution of magnetism in Siberia (Izv. Inst. issl. Sib., 5, 1–8, with 8 maps, 1921).