104 



Dr. Shettle. Investigation 



[May 29, 



the magnetic forces. And, for the purpose of recording the curves 

 so traced out, I covered the surfaces of some bar-magnets with paper, 

 and then ruled lines on them 1-1 6th of an inch apart, longitudinally 

 as well as transversely. 



The squares so formed gave exact co-ordinates for reference. 

 Next, the position of the suspended needle 1-1 6th of an inch long, 

 or l-8th if the curves displayed no sudden curvature, was carefully 

 plotted in each of these squares by means of a hand lens. The ac- 

 cumulation of all these plottings, on careful continuation, gave the 

 curves, as traced from pole to pole, shown in fig. 1. 



The equatorial regions adjacent to the centre of the magnets 

 examined were plotted by dotting down the exact position of the 

 suspended minim magnet in its ultimate state of stable equilibrium; 

 these dots were then assiduously joined by a continuous line. 



In selecting the initial positions of the extraneous curves, equi- 

 distant lines were drawn parallel to the bar, and points were simul- 

 taneously elected on two equidistant lines from which to start equa- 

 torial curves, such as a and b in the fig. 3. 



In every case the points of parallelism of the test needles in these 

 equidistant lines upon being joined by a straight line P P', exhibited 

 obliquity. 



But the principal obliquities are demonstrated by joining the points 

 of contact with the bar of the curves fully traced out from these equi- 

 distant points, such as P P'. 



These curves may be described as those whose tangents are parallel to 

 the axis of the bar N S, at points equidistant from it (fig. 2). 



According to the curvature, changing slowly or rapidly, a different 

 set of needles is employed, which are preserved, ready for use, in 

 glass tubes. 



The work is one requiring great steadiness of manipulation. Some 

 thousands of observations require plotting on a minute scale. Indeed, 

 innumerable curves of great extent and variety have been plotted 

 upon the surface of magnetic bars covered with paper, as seen in fig. 1. 



Not only were these curves traced all over the surface of the 

 magnets, but they were also traced a considerable distance away from 

 them laterally. 



The nature of these extraneous or lateral curves appears graphically 

 hyperbolic. Any given curve thus formed is the position of the magnetic 

 needle traced from point to point, which needle is necessarily under 

 the influence of both the terrestrial and the bar magnetism for the 

 time being, and its direction is a resultant of the various forces bringing 

 it to a state of rest at the various points examined. 



One of the most striking results of the investigation, when the 

 curves are placed very near the centre of the bar, is the general obli- 

 quity of the magnetic equator of the bar (fig. 2). 



