188 KNOTT AND TAN AK AD ATE 



scope and the declinometer coil. Between the Y's, the bar swells out 

 into an oblate ring, through which the magnetometer projects. A 

 semi-circular gro(3ve is cut in front of the ring, S(j that the magnet- 

 mirror can be sighted by the small telesc(jpe. Over the brass ring 

 and surrounding the magnetometer a ring of wood is placed, with 

 three semi-circular grooves cut in it at suitable places. It facilitates 

 observation by cutting away all extraneous light, while at the same 

 time the central groove in combination with the groove on the brass 

 ring gives a clear front to the mirror. The side grooves are needed 

 during night observations to allow the lamp-light to illuminate the 

 scale. On the Y-groove of the bar there arc four stops, two on each 

 side of the centre. The deflection magnet rests in the groove, and 

 the stops are so placed that the two distances of the magnet from the 

 centre are obtauied sim])ly by slipping the magnet along the groove 

 from one stop to the other, witliout having t(3 lift it out. The 

 stops are placed so as to make the ratio of the two distances the best 

 possible, according to the usual rule. 



The instrument is obviously availaltle for use either according 

 to the methcxl of sines or the method of tangents. The former 

 method is the preferable one ; and, in using it, it is necessary to clamp 

 the stao-e to the Y's, and free it from the base of the theodolite. Tiie 

 operations are then conducted exactly as -^vith the Kew Instrument. 

 The temperature of the bar is measured by means of a thermometer 

 placed on the A'-groove outside the further stop. It is advisable to 

 dust with a small brush the surfiices of the magnet and stop just 

 before they are brought together. The cln-onometer time is taken 

 as the final deflection is adjusted. The beginning of the experiment 

 is given by tlie first time record in the vibration experiment, which 

 it was found most convenient to make first. The mean of these 

 times is taken as the time corresponding to the value of the horizontal 

 force as finally deduced. 



