260 Reports on Special Researches 



Referring the mean results in Tables 25 A, 25 B, and 25 C to I. M. S. (see p. 273), 

 we obtain the following values of the quantity (I. M. S. Tiflis): +l'-2 (for declination), 

 +0.00030// (for horizontal intensity), and 0'.7 (for inclination). From S. Savinov's com- 

 parisons in December 1907 at Tiflis (see Table 31 A, p. 270), which unfortunately did not 

 include dechnation observations, we derived (Table B, p. 278) for the value of (I. M. S. 

 Tiflis), +0.00031// (for horizontal intensity) and +0'.l (for inchnation). The agree- 

 ment in the two independently-derived values (+0.00030// and +0.00031//) for the 

 //-difference is very satisfactory. The two values ( 0'.7 and +0'.l) for the /-difference, 

 however, do not agree as well as is desirable. 



Pending the receipt of the Observatory data for our 1909 comparisons, which were obtained 

 under better conditions than in 190S, ive defer giving our final results at Tiflis. 



No. 26.-TOKIO OBSERVATORY, JAPAN. 



Comparisons were secured with the standard magnetic instruments of the Central 

 Meteorological Observatory at stations in the grounds of the Tokio University during 

 August and September 1906, when the magnetic survey vessel, Galilee, was at Yokohama. 

 The instruments used by the members of the Galilee party were C. & G. S.^ magnetometer 

 No. 36 and C. & G. S. dip circles No. 35 (with needles 2 of circle 35 and 5 of circle 163) and 

 Dover No. 178 (with needles 1, 2, 5, and 6). The observations with the Tokio instruments 

 were made by Messrs. S. Nakamura and Y. Oislii, under the direction of Professor Tanaka- 

 date. The Tokio instrument used for declination and horizontal intensity was of the pat- 

 tern of magnetometer devised by Professor Tanakadate and descril^ed in the Journal of 

 Science, Imperial University, Japan (Vol. II, Part III). For inclination, the Tokio stand- 

 ard, a Casella dip circle (No. 17?), and a Wild earth inductor were used. 



All the Tokio observations were obtained in the absolute house of the Observatory, 

 while most of the C. I. W. observations were made at the main C. I. W. station about 30 

 feet north of the absolute house; some dip observations also were secured by the Galilee 

 party at a secondary station located about 30 feet west of the main C. I. W. station. The 

 stations outside the absolute house were selected after consultation with Professor Tana- 

 kadate; the station-differences were not determined, as it was understood that they were 

 negligible. The observations by the two observing parties were not always strictly simul- 

 taneous, and it was necessary, therefore, in reducing the two series to equivalent local mean 

 times, to make use of copies of the declination magnetograms and of the published hourly 

 results of hortizonal and of vertical intensity. 



Subsequent to the work of August 15 and 16, 1906, the C. & G. S. magnetometer 

 No. 36 suffered to some extent by having been submerged in its box during the accident 

 which befell the Galilee in Yokohama Baj^ owing to a typhoon. The work of September 3, 

 1906, was carried out after the instrument had been thoroughly cleaned. 



The azimuth of mark (lightning-conductor on Chemical Laboratory) was, as supplied 

 by Professor Tanakadate, 179 55'.4 from the absolute house, and 179 54'.6 from the main 

 C. I. W. station. 



In view of the non-exchange of stations, lack of simultaneity of observations, unequal 

 number of determinations by the two parties, and the uncertainties involved in the reduc- 

 tions to the same times and to the C. I. W. standards, the results of the comparisons should 

 be regarded as wholly tentative ones. It is hoped in the near future to obtain better 

 determinations of the relations between the Tokio and the C. I. W. standards. 



'C. & G. S. stands for Uuited States Coast and Geodetic Survey. 



