2 INTRODUCTORY NOTE. 



been so light, compact, and easily portable. The asbestos houses were fairly satisfactory, but had 

 some disadvantages. 



l*gtaUaKa*. The instruments were erected on a strong firm wood bench (> p. 3), 8 feet 5 inches 

 in length and 1 foot 6 inches in breadth, and supported at one end by a drain pipe, 1'6 feet in diameter, 

 sunk into the frozen ground, and at the other end by a thick pillar of wood sunk in the same manner. 

 The thickness of the wood slab forming the bench was :i inches, and it was 2-6 feet above the floor 

 of the house. The bench was carefully erected in the magnetic Meridian, the magnetometer and 

 Declination magnet being used for finding the Meridian, and the instruments fixed upon it in the 

 following order : 



Magnetic North (S.S.E. true) extremity the recording cylinder and clock apparatus, nearest the 

 cylinder the declinometer, then the Horizontal-Force instrument, and at the magnetic South extremity 

 the Vertical-Force instrument. 



The suspension fibre of the declinometer was the original one employed at Kew ; it was used 

 throughout the two years and returned to England with the instrument. 



On examining the Horizontal-Force fibre originally supplied with the instrument it was found to be 

 broken ; this unfortunately happened to be the stoutest fibre supplied, and it became necessary to replace 

 it by a succession of fibres which gave the magnet an unnecessary and rather troublesome degree of 

 sensitiveness. 



The magnets for the declinometer and Horizontal-Force instrument consist of well-hardened laminar 

 pieces of watch-spring steel, 25 mm. in length and weighing about T5 gramme. A light aluminium frame 

 supports the mirror and the magnet, and this is hung by means of a double hook on a small cross- 

 piece attached to the bottom of the quartz-fibre suspension. 



The Vertical-Force instrument is a modification of the Lloyd balance, and as it was only completed 

 shortly before leaving England, very little was known of its behaviour even .at Potsdam. This 

 instrument was a source of constant trouble. The needle was balanced for a dip of about 70 N. 

 The magnetic dip at Winter Quarters being nearly 85 S., the pull on the S. end of the needle 

 could not be overcome by the small weights and auxiliary magnets supplied for the purpose, and 

 therefore additional weights had to be added to the N. end, which increased the temperature co- 

 efficient of the balance. 



The chief feature of the recording apparatus is that all three elements, base lines, and a temperature 

 curve are on the same photogram for the day. On disturbed days and they were frequent and in 

 the Summer when the movements of the magnets are large the result was a considerable confusion of 

 the curves. 



From May, 1902, until January, 1904, the declinometer was never interfered with, to the knowledge 

 of the observer, nor its zero mirror altered. 



During the first year the Horizontal-Force instrument was two or three times found to be out of adjust- 

 ment and altered, but remained untouched after April 1, 1903, while the Vertical-Force instrument was 

 altered from time to time during both years. The method of determining sensitiveness was by deflecting 

 the suspended magnets with one of the unifilar collimating magnets at certain known distances, and then 

 carefully finding the moment of the deflecting magnet by a set of absolute observations. 



Sotitine. The general routine was usually as follows : The Observatory was entered nt between 11 a.m. and noon each 

 day, the light-shutter of the magnetograph closed, and the time of doing so noted by means of a chronometer watch. The 

 thermometer inserted in the Vertical-Force instrument was then read. 



After changing the paper on the recording cylinder, filling and trimming the lamps, the thermometer was again re .id, the 

 light-shutter dropped, and the time of doing so noted as before by means of the chronometer. The whole operation 

 occupied about 30 minutes, and times of stopping, starting, temperatures, and error of watch on mean time were entered in 

 a note book. 



Temperature of House. During the first year the walls only of the Variation House were banked with 

 snow, and a large brass heating lamp was kept burning within, so as to maintain as uniform a temperature 

 as possible. This lamp was frequently a source of danger and inconvenience of an aggravating nature, 

 and required constant watching. Unless there was a draught underneath the lamp it emitted dense 

 smoke which on calm days filled the room and extinguished the small oil lamp which provided the beam of 



