118 



three points in line, point the telescope to the mark at the North, read the needle at 

 both ends, and, in order to eliminate all errors of eccentricity of the graduation or 

 needle and thereby arrive at correct results, the instrument should be reversed on its 

 vertical axis and the telescope again pointed to the mark at the north end as before and 

 another reading taken from both ends of the needle. 



The average of the four results will be the combination of both the index and 

 magnetic errors and that of the declination of the needle the latter bein^; ili< 

 the magnetic meridian makes with the astronomical meridian at the place of ol>M 

 tion. Naturally, then, in a theoretically perfect instrument the reading of a need I 

 place where East and West declinations join, would be zero when the cross win-s biseet 

 the meridian mark. However, as in the field use no instrument will retain its line 

 qualities and adjustments, it will be best not to depend on them, but to determine the 

 index error and constant of the needle combined simultaneously, from time to time, by 

 an observation on a meridian as above described, in order to apply it as a correction in 

 precfse work. In instruments fitted with a variation plate permitting the declination 

 for any particular locality to be set off, the combined errors of the instrument and 

 needle can at once be added to or subtracted from the declination, as the case may be. 

 But, as these constants differ in every instrument it furnishes the explanation to the 

 question propounded in the title of this paper: Why do different magnetic needles not 

 always point in the same direction, though observed at the same time and place? 

 Unless these constants have, been previously determined and applied in every case. 



There are other reasons, such as the daily variations of a needle, which, according 

 to temperature, latitude and season of the year, may amount to from six to sixteen 

 minutes alone, showing why reliance on the reading of a magnetic needle can be placed 

 only within certain limits and that it should be depended on only for general direction 

 as required in filling in details. For exact work meridian lines run by solar or stellar 

 observations should be depended on exclusively, to which end the modern solar attach- 

 ments give the fullest satisfaction. 



In conclusion, we would say that such tests on a meridian mark, as recommended 

 above, should be repeated often, and that in order to facilitate the work such meridian 

 marks should be located at some convenient place so as to be within easy reach of the 

 surveyor at all times. 



In the selection of such a place care should be taken that there is no iron in the 

 vicinity and that there are no electric wires, particularly those used for arc lighting or 

 the conveyance of power within a radius of three to six hundred yards, or else a needle 

 is apt to be affected by these currents as would be the needle of a galvanometer. Such 

 a place, if properly selected and provided with permanent meridian marks, could at the 

 same time be made available for verifying the telescope and level adjustments of the 

 transit so necessary to its proper use. For the benefit of surveyors, the City of Boston, 

 many years ago, provided such meridian marks on stones deeply set into the ground on 

 Boston Common. Unfortunately, electric wires now pass all around these grounds, and 

 even traverse them, so that reliance can no longer be placed on observations there 

 made. 



The Berger Short Focus Lens Attachment. 



A very valuable addition to the engineer's out- 

 fit is found in the short focus lens attachment 

 which has been brought out. The contrivance is 

 simple, but, like many simple devices, is very 

 effective in overcoming a practical dittieulty. 

 Probably every engineer has been annoyed by being 

 obliged to sight a point a little too near for the 

 telescope to focus. Most transit telescopes will not 

 focus on a .point much nearer than 6 or 6 feet 

 (levels not nearer than 7 or 8 feet) away from the 

 instrument, while it is frequently necessary to si.uht 

 a point on the ground nearly under the transit, at 

 a distance which is usually less than that. 



In mine surveying as well as inside of factory 

 buildings, one frequently needs to sight a point 

 overhead or on the walls and very near the transit. 

 Ordinarily the only way out of the difficulty is to focus as nearly as possible and do the 

 rest by a guess. As a further instance, one often finds in leveling, that it will be n 



