TACHEOMETER SURVEYING WITH AN ORDINARY THEODOLITE. V. 



one-hundredth of a foot at the end ; but in places the differ- 

 ence in level amounts to 1*65 feet, and the distance differs 

 by 26 links. As the errors give and take the results are 

 quite near enough for preliminary surveys. 



On five miles forty-six chains of trial survey that was 

 staked out with tacheometer and afterwards chained and 

 levelled in the ordinary way, the fiual result gave the 

 difference in distance of 84 links, and the difference in the 

 reduced level of 7*30 feet. The number of traverse lines 

 was 171 and their average length 2*6 chains. The short 

 lines were unavoidable on account of the rough nature of 

 the country, as the line staked ran under a steep moun- 

 tainous slope covered with a dense undergrowth. Upon 

 re-levelling, it was found that the errors were mostly in 

 the steep slopes, and it was afterwards found advisable to 

 repeat the angles and to exercise great care in setting 

 them on the vertical arc of the instrument. The table was 

 originally calculated to the nearest minute only, but this 

 was not accurate enough, so a table of angles to the nearest 

 second was substituted. On the cross sections the same 

 accuracy is not necessary. 



The theodolite should have the bubble attached to the 

 vertical arc and not on the telescope. The theodolite used 

 had the bubble removed from the telescope and fixed on 

 the arc, and being much longer and more sensitive than 

 the bubble usually attached to the arc, the result was very 

 satisfactory, and the cost of removing the bubble was 25s. 

 So as to enable the staff holder to hold the staff truly vertical, 

 a longitudinal bubble was attached, this being frequently 

 adjusted with a plumb bob. For ordinary levelling it has 

 been found that the bubble on the staff gives greater 

 accuracy and more expeditious levelling than waving the 

 staff. Some difficulty was experienced in reading the staff 

 at a greater distance than four chains, and it would be 



