115 



point in the needle cap, may not coincide with its magnetic axis, which lies in the plane 

 passing vertically through its poles and in which the magnetic meridian is contained. 



However, to get a complete understanding of the matter, we must go back to the 

 ingot from which the steel of the needle was produced. If we remove a portion of the 

 outer crust of a steel ingot or other casting and carefully examine the surface laid bare 

 with a magnifying glass, we shall find that what seemed to the naked eye as a solid mass 

 is now a spongy one, with numberless small and large cavities or blow holes. It is only 

 by forging and rolling into bars and sheets, at the mill, that the metal becomes at all 

 homogeneous. Some of the cavities not being welded, are elongated in the direction of 

 the rolling, and form veins, flaws or blisters according to size, thus imparting to the bar 

 or sheet a structural grain or fiber, which, while always traceable in poor steel, can only 

 be detected in degree when the ingot has been converted into fine steel. In the absence 

 of information to the contrary it is therefore but natural for the writer to believe that 

 when a needle is being magnetized its resulting magnetic axis will in all likelihood tend 

 to run in the direction of the general trend of this grain without strict regard to the 

 geometrical axis, with which it may then be at an angle. The above divergence of these 

 two axes may, therefore, be considered as one cause of the observed fact that two 

 needles, though of the same size and shape, other things being equal, may not read exactly 

 the same when observed at the same place and time. 



If, on the other hand, these two axes (magnetic and geometrical) do coincide, there 

 will be no deviation, and two or any number of needles will read alike under the con- 

 ditions above mentioned, if tested in the same compass. 



Notwithstanding the fact that it is, in practice, very difficult to attain the desired 

 result on account of the difficulty of tracing the grain, the writer would and does insist 

 that the steel used in the construction of needles for surveying instruments shall not be 

 forged, but shall be cut from fine shear steel, in the direction in which it was rolled, and 

 be treated in such a manner as to ensure as near a parallelism of the structural grain 

 with the geometrical axis as possible. Yet in spite of the utmost care and skill con- 

 sequent minor magnetic poles are probably present, and there is almost certain to be a 

 deviation of these two axes, and this deviation will vary in magnitude with the quality of 

 the material and grain, the design, width and length of the needle, and the mechani- 

 cal skill with which its outlines in azimuth have been made, as also with the degree of 

 accuracy with which the point of suspension has been located in the geometrical axis. 



While it may seem that just here the instrument maker's real work should begin 

 in the testing and adjusting of these axes by grinding off a little from one side or the 

 other, as the case may be, to insure a coincidence of the two axes, he, as a rule, will 

 stop here either from lack of understanding or from want of necessary apparatus and a 

 secluded space free from iron and provided with well defined meridian marks estab- 

 lished by means of collimators or natural objects. 



His chief reason for not doing this, however, may be said to be the fact that neither 

 he nor the surveyor wish to bear the added expense. 



For this and other reasons the scientist, not wishing to depend upon mechanical 

 skill for good results, adopts the method of reversion, by means of which errors from 

 this source can be entirely eliminated. 



Therefore, the needle used in his investigations (See Fig. A) can be reversed by 

 simply turning the needle upside down (the cap having been changed from the top to 

 the bottom side of the needle) on the center bearing point, so that be can use the mean 

 of the readings of each end in both the direct and reverse position as the result 

 sought. 



If the surveyor is desirous of making some such test he can do so readily, not by 

 changing the center cap as above noted, for he cannot do so with American instru- 

 ments, as the center cap is firmly fixed to the needle, but, by first taking a careful read- 

 ing of both ends of the needle, he can, by means of a strong magnet, reverse the poles 

 of the needle,* and, after rebalancing, read again the two ends, thus obtaining a read- 

 ing correct in a measure and approximately finding the constant of his needle, which 

 he can use when needed for important work. 



The constant of the needle is therefore the angle that the magnetic axis makes with 

 the geometrical axis. 



If a needle has aluminum extension arms, as in galvanometers, Fig. 9, or a circle 

 attached as in the case of prismatic compasses, Fig. 5, or if it is provided with verniers 

 as in some compasses, Fig. 6, then the constant of the needle, or the index error is the 



*Accidental changing of the poles by carrying the instrument on an electric car near the motor has been 

 noticed by the writer, and has explained what, at first, seemed to be very queer behavior on the part of the 

 needle. 



