OF ECCENTRICITY AND GRADUATION. 33 



Avith two collimator images as Avith one image bisected by a wire. In either 

 form the circle should be provided with microscopes, which are more 

 expeditiously read, and less fatiguing to the eyes, than verniers. 



The examples which have been discussed in the preceding pages are 

 sufficient to show that, under favorable circumstances, observations can be 

 made with the sextant leaving very little to be desired as regards accu- 

 racy. In practice, however, such precision is not always, perhaps not 

 commonly, attained. The most insidious source of error is au unstable 

 condition of the eccentricity — a fault clearly traceable to defective con- 

 struction when due attention has been paid to the care and preservation 

 of the sextant. A judicious observer always endeavors to distribute his 

 observations so as to neutralize the unknown errors of his instrument as 

 nearly as possible; but variations in the eccentricity, which may occur at 

 any moment, cannot be evaded by this means. If the conical axis is so 

 improperly fitted as to be circumferentially supported only at the smaller 

 end, its position will probably be maintained by friction, and the viscosity 

 of the wax-like lubricant used for this bearing, until some extraneous 

 force is applied, when displacement into a new position of temporary 

 quiescence may be expected to ensue. Such a movement, to the 

 extent of one-thousandth of an inch in a sextant of seven inches radius, 

 may produce errors of 50" and upward. Any unavoidable defect in 

 fitting should evidently subsist in the direction opposite to that here sup- 

 posed, and possibly some changes in the usual dimensions and materials 

 of construction might be advantageous. But whatever the requisite 

 alteration in existing practice may be, its discovery and adoption can 

 safely be intrusted to the instrument-makers if a sufficient inducement to 

 persevere in the search for improvement is oflTered to them. The 

 mechanician who expends time and money in striving after a perfection 

 which observers do not demand, or appreciate, unwisely impairs his 

 ability to compete with his rivals. When sextants are so generally and 

 adequately tested that the reputation of each maker rests on the actual 

 merits of his work, a remedy for the evils of injudicious design and 

 inferior workmanship will soon be found ; until that time it cannot rea- 

 sonably be expected. 



