106 MAGNETIC OBSERVATIONS 



fact a better setting will be obtained with the needle slightly in vibration 

 and bisecting the arc. The bit of steel used must, of course, be removed to a 

 safe distance before the final setting. Record the reading and quadrant in 

 which it falls and note the time by the pocket watch used in the sun observa- 

 tions. Do similarly with the south end. Next slightly disturb needle, in 

 same manner as before, and repeat pointings in reverse order, i. e., first south 

 end and then north end. (d) Shift horizontal circle until the north end of 

 needle instead of cutting the zero, cu.ts the mark 5 E. of N. and carry out the 

 same operation as for c. (e) Shift horizontal circle until the north end of 

 needle now cuts the 5 W. of N". mark and do as before. (/) Shift horizontal 

 circle so that north end of needle again cuts the zero and do as in c. Record 

 the time of ending, clamp needle and again make two pointings on the mark as 

 at the beginning. 



2. With Compass Needle of Theodolite. The method already described 

 can readily be adapted to this needle. The pointings on the mark in this case 

 will be made with the telescope and, as the horizontal circle is graduated con- 

 tinuously from to 360, it will be unnecessary to note the quadrant in which 

 the reading falls; however, both verniers should be read. 



If the work could be so arranged that the declination would be obtained 

 with one needle in the morning, say about 8 to 9 o'clock, and with the other 

 needle from 1 to 2 o'clock p. m., the mean of the two^results would be almost 

 entirely free of correction due to diurnal variation of the earth's magnetism. 

 The declination reaches its mean value between about 10 to 11 a. m. and 

 5 to 6 p. m., local mean time, and its extreme elongation at about the times 

 above mentioned. 



C. DIP OBSERVATIONS. 



(The compass needle must be removed from instrument and tent before 

 observations are made.) 



The regular dip observations following the method already shown and 

 as prescribed by specimen, should be made as already stated, with the Dover 

 needles N"os. 3 and 4. These needles should invariably be reversed before 

 making the observations and again at the conclusion of the half sets. The 

 time to the nearest minute of beginning and ending for each needle should be 

 recorded. The order, after the magnetic prime vertical with one needle has 

 been determined, would be as circumstances condition: Either (a) Dip ob- 

 servations with needle 3, with say, A down, then observations with needle 4, 



