370 



Special Reports 



Brown into Manchuria, using the same circle and needles Nos. 2 and 5 but replacing 

 Nos. 1 and 6 by two others. The corrections on adopted standard for the needles used 

 during both expeditions for inclination +62. 2 were: in Mongolia during September and 

 December 1915, +0'.6 for No. 2 and +5'.2 for No. 5; in Manchuria during September 

 1916, +0'.4 for No. 2 and +0'.2 for No. 5. The corrections for September and December 

 1915 are taken from the curves adopted from the consideration of a large number of 

 stations, while those for Manchuria are from an analysis based on two stations, these being 

 the only ones at which observations were made at inclination values suitable for compari- 

 son. The large correction for needle No. 5 which was obtained for the Mongolian obser- 

 vations applied equally well in September and December 1915, but no evidence for the 

 necessity of such a large correction could be found from the observations in September 

 1916, indicating that, whatever the cause might have been, it persisted for a few months 

 only. 



Fio. 10. Observed and Computed Correction-Curves for Needle No. 5 of Circle No. 177. 



One further cause for the erratic behavior of needles has recently come to our 

 attention. Dip circle No. 177, which has been a superior instrument, was equipped with 

 four new needles for use in the campaign of 1919 to 1920 through central Africa. These 

 needles gave results which differed among themselves by unusually large amounts, 

 despite the observer's conscientious efforts to bring them into harmony by careful atten- 

 tion to the pivots and repeated readings, indicating minute pivot-defects. Needle No. 

 16X, as the work progressed, showed most unusual and erratic behavior; it was ulti- 

 mately discovered by the observer that the pivot of this needle had become so loose in 

 its sheath that it could be easily withdrawn. Obviously cleaning such a loose pivot by 

 a rotary motion in pith, tissue'paper, or chamois, as is the practice, might rotate it slightly, 

 thus altering the balance of the needle and causing considerable variations in the results. 

 It is not supposed that such conditions have often occurred, but they may be responsible 

 for some difficulties otherwise impossible to explain. In this case it may have been due 

 in part to extreme daily range in temperature, expanding unequally the steel pivot and 

 its brass sheath. 



