llO Transactions of tlie Boyal Society of So2ith Africa. 
i.e. the greatest change occurring, whether from higher to lower or lower 
to higher, in one day which has taken place in any assigned month of the 
period dealt with. For example, '409 inch is the greatest change of 
pressm-e which has taken place in one day in January at Cape Town. 
In this Table the various stations are not exactly comparable one with 
another. For, obviously, the longer the record at any place the greater is 
the chance of getting large values for the maximum variability. Taking 
the Table as it stands, however. East London, of the South African stations, 
has had the largest individual value of the variability, namely '735 inch in 
August ; Cape Town coming next with -711 inch, also in August. Kim- 
berley, and more especially Daressalam, happen far short of such a value. 
Of the Australian stations Adelaide and Sydney are comparable with the 
coast stations of South Africa; while Alice Springs, in the interior of 
Australia, is pretty much the same as Kimberley in the interior of South 
Africa. 
Of the 13 stations in the Southern Hemisphere, Staten Island is 
easily first, in spite of its short record, in the matter of barometric 
oscillation. 
Table 5, perhaps, gives more useful information than the previous 
Table. It gives the average values of the monthly maxima of variability 
for all the stations that have a record long enough to be worth considera- 
tion in this particular; i.e. it shows the average for each month of the 
maximum variability which has occurred in that month in each year. In 
other words, take any station N and any month M, and let the maximum 
variability in that month in the first year be a^, in the second year ttg, . . . 
and in the nib. year Then the mean value of all the maxima will be — 
— (a, 4- a^ a., 4- ... + a„). 
n ^ 
Thus at Cape Town, in January, the mean value of the maximum varia- 
bility is -277 inch, the greatest value of all the individual maxima in that 
month being '409 inch (in 1903), while the least value of all the maxima is 
•189 inch (in 1904). It is curious that if we pick out the greatest variability 
in each month for some extended period, the average of them all is prac- 
tically three times the constant. For example, the annual mean value of 
the maximum variability at Cape Town is (Table 5) -332 inch, while the 
constant is (Table 3) '108 inch. A point worth remark is that whereas the 
monthly constants are on the whole greater at East London than at 
Durban, yet Durban experiences the greater maxima, particularly in the 
summer months. 
Tables 6 and 7 give the greatest and least values of the monthly and 
yearly mean variability. For example, in determining the monthly mean 
constant -089 inch at Cape Town for January we meet with -105 inch in 
