236 Manipulations ^of the Dipping Compass. 



the centre of the graduated circle, by the same rotation which 

 brings it to the true dip. To restore the pivots to their place, two 

 Ys and two incHned planes* raised by a lever engage them at 

 their sides and at their ends, adjust them laterally and longitudi- 

 nally, raise them just clear of the agate edges, and, by a reverse 

 motion, lower and deposit them, and with them the needle in 

 their proper places. Now the most objectionable anomaly has 

 been, that when the same needle has been once adjusted by the 

 Vs, read, and again simply raised by the Vs, and lowered appa- 

 rently in the same place, and read again, there would be a dis- 

 cordance in the readings, often amounting to five minutes of a 

 degree. I felt the full weight of these anomalies in my first at- 

 tempts to determine the dip, and found to my mortification, that 

 the mean results of the usual eight reversals and sixteen readings, 

 with two different needles, would sometimes differ as much as 

 six minutes of a degree. 



I finally reflected that if the pivots could always be made to 

 re-sit on exactly the same points, the readings must always agree, 

 and that the apparent anomalies must arise from slight iniperfec- 

 tions of form, and slight and imperceptible, but really essential, 

 changes of position of the points of support to the pivots. With 

 this view I endeavored by all possible means, so to use the appa- 

 ratus as to bring the bearings at the same points, and especially 

 that the pivots should perform their rotations on the same trans- 

 verse section ; or, in other words, that the same ring or circle of 

 their circumference, should always rest on the agate edges. To 

 accomplish this, the following points received attention. 



1. The Vs and inclined planes, which raise and fix the needle 

 to its place, were so adjusted as to allow of no shake or lateral 

 motion of their own.f 



2. The distance between each agate edge and the inclined 

 plane, opposed to the end of the pivot operating as an end check, 

 was made as nearly equal as possible. 



3. Although the pivots had some "end chase" or longitudinal 

 freedom between the end checks yet in use, each pivot was al- 

 ways slid over against one particular end check by means of a 



* These inclined pianes slope up opposite the ends of the pivots, and are really 

 two parts of a V widely separated. 



t I have lately examined one of Gambey's instruments, which was decidedly 

 faulty in this particular. 



