10 A. TANAKADATE. 



made by Ciisella, these are Nos. ÖG13, 5G14, and 0015; after 

 1894 tliey are nuni leered as 1, 2, 3, respectively and are put 

 together witli the magnetometers of corresponding numl)ers. In 

 1893 one dip circle (Casella No. 4365) was borrowed from the 

 Hydrographie Bureau for the East Party and another (Dover No. 

 88) from the Science College for the West Party. 



For reversing the magnetisation of needles a pair of coils 

 each wound with insulated wires of three difterent diameters ; 

 the thinnest being put innermost, was made for each set of in- 

 struments. They were put in a closely fitting box which acted 

 also as sliding guide in introducing the needle safely into the 

 narrow cores. The usual bar magnets were only carried to 

 provide for the case of the failure of dry cells. 



The declinometers were electromagnetic, the same as those 

 used in the survey of 1887, the only difference being in tlie 

 construction of the mirror magnetometer already described. 



Simultaneous ol)servations were made with those instruments 

 at various times as is seen in the Complete List of Observa- 

 tions, and differences were found. They arise from the errors 

 in the determinations of the constants, from the errors of gradua- 

 tion and the eccentricity of the circles, from the deviation from 

 circular form in the axil of dip needle besides the accidental 

 errors of observations. The difference will depend upon the 

 local values of magnetic elements and different sets of observers 

 to certain extent, hence we have not applied instrumental 

 corrections in any of the results, the constants of the magne- 

 tometers were however determined at various intervals during 

 the survey. 



The distribution of these instruments among the several 

 parties were as follows : — 



