294 Transactions of the Boyal Society of South Africa. 
Elizabeth Harbour Board, and worked up the directions for each hour. 
Probably if he had done so he would have found the same rule as prevails 
at various other coast places, i.e., that the winds belong to three super- 
imposed systems : (1) a more or less strong counter-clockwise diurnal 
rotation depending to some extent upon the season of the year ; (2) pre- 
vailing winds (depending upon the pressure gradient) up or down the 
coast ; and (3) a slight land-and- sea-breeze effect which at once perturbs 
the diurnal rotation and causes a deviation land- or sea-wards of the 
prevailing gradient winds. No study of South African winds promises 
to be quite complete unless it comprises at least four equidistant observa- 
tions a day. When the observations are only made at two equidistant 
hours {e.g., VIII. and XX.) they can only be of use as extending the 
continuous observations made at some adjacent observatory. 
In this connection the winds of Delagoa Bay are of interest. The 
information given below is partly summarised and partly computed 
from the important second annual report of the Observatorio Campos 
Eodrigues. 
The average velocities for the year 1910 are : — 
TABLE 1. 
Mean Houely Yelocities of the Wind for each Month at 
LouRENCo Marques. 
Kilometers per Hour. 
Miles per Hour. 
20-94 
13-0 
21-05 
13-2 
March 
22-03 
13-8 
April 
20-67 
13-1 
May 
21-01 
13-0 
21-92 
13-6 
July 
20-91 
13-0 
23-94 
14-9 
26-83 
16-6 
October 
25-91 
16-1 
November 
27-19 
16-9 
2594 
16-1 
23-19 
14-4 
For the year, therefore, this gives a mean velocity only 72 per cent, of 
that of the wind of East London. 
