BY Wivks TULEY? BSQA RIAL EARG.S., ETC. 135 
measurements and the corrections for temperature were readily 
borne in mind, and, whilst the measurement was being carried 
out afforded a most efficient check, any divergence being at once 
noticed, and as it indicated that the thermic conditions of the 
tape and the thermometers did not agree, one or more additional 
measurements were taken until the opposing differences were 
reconciled. Whilst the temperature was steady the micrometers 
and thermometers worked together most harmoniously—rapid 
fluctuations on the contrary produced conflicting readings. The 
advantage of having such a check cannot be over-rated. It gave 
an additional value to the work. The confidence which it 
engendered reacted on the observers and enabled them to give 
a more undivided attention to the details than was possible 
under other circumstances. 
I give in appendix B the computed lengths of each separate 
measurement and the difference between the mean of each 
section. The difference between the mean of the three measure- 
ments of each tape was 1845144 inches, which is not quite 
# of an inch. The adopted lengths of the tapes were as 
follow :— 
Tape A - - - 99°9980561 feet 
Tape B - - - 99:9976211 ,, 
Difference - - (0004350 _ ,, 
This fractional part of a foot is nearly equivalent to 73, of an 
inch. This difference in length between the two tapes, were it 
constant throughout the entire distance from end to end of the 
base, would have amounted to 1°8 inches; but as the actual 
difference between the measurement by each tape was not quite 
+2; of an inch, it proves that about an inch and a half of the 
theoretic difference was absorbed by errors of observation or 
other counterbalancing cause. The measurement by each tape 
has equal weight, and the mean of the two is only affected by 
the above absorption to the extent of ‘7 of an inch. The above 
