Captain Kater/s floating collimator , &c. 311 
where these letters do not occur, the observations were made 
b y a grosser wire. 
The mean index corrections for this latter wire, as deter- 
mined by the collimator, were as follow : 
To Aug. 22 inclusive, . Index correct. CE — 56",30 
South of zenith \ W -f- 54,46 
^ CE — 57,44 
From Aug. 25 to 29, tw + 55,92 
From Aug. 29 to Oct. 7, face W, . . . . + 59,9 o 
From Oct. 7 to Nov. 10, face E, . . . . — 59, 11 
From Nov. 10 to Dec. 3, face W, . ... -f- 58,62 
By the fine wire : 
From Oct. 15 to Nov. 10, face E, .... — 15,10 
From Nov. 10 to Dec. 3, face W, . . . . + 15,01 
The interval between the wires is 1' 13", 92 ; and therefore 
previously to Oct. 15, the index correction for the fine wire 
is found by subtracting the index correction for the thick 
wire from the interval 73", 92. From Oct. 15 the index cor- 
rection for the fine wire was found directly. 
Explanation of, and remarks on Table II. 
This table gives the inclination of the line of collimation of 
the collimator to the horizon, as observed on different days. 
It is not necessary that the collimation of the collimator 
should remain permanent from day to day ; but when it is 
found to do so, it stamps an additional value on the instru- 
ment. It is only necessary that derangement should not 
take place in carrying it from one box to the other. This 
might be feared, when we consider how small a change in 
the relative position of the parts of the instrument would dis- 
