346 
Transactiojis of the Boyal Society of South Africa. 
It appears from this that the greater the difference between the ranges 
of temperature for any month given in Tables 3 and 4, then the greater 
the ratio of a to /3 on the whole for both of the first two harmonic terms. 
No great faith can be placed upon the result for the second harmonic 
term, though. At the same time the ratio is suggestive, and should it be 
confirmed by a longer series of observations, we should have to conclude 
that some portion of the second harmonic term of pressure is of local 
origin. 
The phase times of the harmonic terms of temperature, given in 
Tables 5 and 6, are also practically unaffected whether the range of 
temperature be greater or less. The amplitudes, however, are rather 
more modified, than those of pressure. The following are the a/fi 
ratios : — 
_»2_ 
alii 
a//5 
a//3 
o 
6-1 
1-295 
1-246 
1-431 
Feb 
6-1 
1-261 
1-305 
1-322 
Mar 
5-4 
1-270 
1-196 
1-525 
6-5 
1-333 
1-197 
1-961 
May 
7-5 
1-409 
1-386 
1-856 
3-7 
1-149 
1-063 
1-611 
5-5 
1-269 
1-171 
1-004 
Aug 
3-1 
1134 
1-126 
1-257 
Sept 
5-5 
1-218 
1-271 
1-400 
Oct 
4-4 
1-198 
1-195 
1-171 
Nov 
3-9 
1-160 
•982 
1-597 
Dec 
3-1 
1-143 
1-112 
1-118 
5-1 
1-230 
1-184 
1-386 
The ratios evidently depend fairly closely upon the value of a — /3, the 
first term more so 
than the other two do. A 
summary is :— 
No. of 
M-m 
i 
Cases. 
a//3 
a//3 
alii 
7 
o 
6-2 
1-294 
1-253 
1-500 
5 
3-8 
1-157 
1 
1-096 
1-351 
