14 KIGIITH REPORT 1833. 



tached to the inside of the sliding frame, and exactly behind the 

 centre of the white circle, is a vernier, nea'tly in contact with 

 the face of the staff, which divides the hundredth of a foot into 

 five parts, of 20 ten-thousandths each, so that the observation is 

 read off and recorded to four decimal places. The white circular 

 spot, and the angular spaces between the lines, may be bisected. by 

 the horizontal wire of the telescope, with great exactness. In 

 favourable weather, I have usually found the average error, or 

 the difference of a single reading from the mean of the number 

 taken to be about ^ ju'^'^ ^^ ^^^ inch, on a distance of 88 yards, 

 or about a quarter of a second of angle. (See Wood-cut at the 

 end of this paper.) 



When the vane was raised so near the top of the staff as to be 

 out of the reach of the hand, the adjusting screw was worked by 

 a long fork of stout wire thrust into holes made in the milled 

 head to receive it. A groove made in the upper part of the 

 staff receives the fork when it is not in use. 



In leveling, I proceeded regularly in the following manner. 

 Two equal distances, usually of 4 chains or 88 yards each, ha- 

 ving been measured forwards from the last station, the level was 

 placed at the end of the first distance, and, at the second, a 

 strong wooden peg driven firmly into the ground, for the fore 

 station, the level being exactly midway between the stations. 

 When, (as happened in a very few instances,) I was prevented 

 from making the fore and back distances equal, compensating 

 unequal distances were immediately afterwards taken, so that 

 the sums of the two sets of distances were kept equal through- 

 out. The staff being held vertically on the back station peg, by 

 the means before described, and the first observation taken, the 

 height was read off and written down by the assistant in a rough 

 minute-book wliich he carried for the purpose. The vane was 

 then purposely thrown out, by turning back the screw, the level 

 re-adjusted, and a second reading taken. If these readings 

 agreed within 20 lO.OOOths (about ^^th of an inch), the staff 

 was brought forward to me, when 1 read off and inserted the 

 last reading, according to both scales, in separate columns of my 

 book ; the mean of both readings was also inserted in a third 

 column, after my assistant and myself had called over and com- 

 pared the last reading. The assistant then read off and called 

 to me the last reading from the large scale, as a check on what 

 I had entered in my book. The needle bearing and distance in 

 links, being also inserted in their respective columns, completed 

 the back observation. The process in taking the fore observation 

 was the same, except that instead of having the staff brought to 

 me to be read, 1 had then to carrv forward mv level to the staff. 



