162 Transactions of the Royal Society of SontJi Africa. 
maximum much later, this lag being governed by the nature of the land 
surface, being later over sandy plains and earlier over forests. 
It will thus be seen that there is a constant state of instability between 
the air over the sea and that over the land. At sunrise the air above the 
land has cooled much below that over the sea, and the tendency is for the 
prevailing wind to shift a point or two so as to blow more seawards. As 
the day advances the temperature of the air in general increases, and at 
the solar noon the maximum is reached over the sea. By this time the 
air temperature over the land and sea are on an equal footing. After the 
solar noon, however, unstable conditions set in again ; the air temperature 
over the sea begins to fall, but that over the land continues to rise, and in 
due course the prevailing wind shifts so as to blow more shorewards than 
in the morning. 
These are the broad lines of argument to apply to any investigation of 
this kind. 
The present investigation was the result of pure accident on the part 
of the author. He was about to ascertain the mean direction of wind at 
various stations in South Africa, and for this purpose utilised the Eeports 
of the Meteorological Commission for 1876 to 1879 inclusive. Of all the 
Eeports of the Commission these are the only four which give bi-daily 
observations of wind directions, and these had to suffice, for the present at 
any rate. 
Port Elizabeth was the first station dealt with, and it was soon seen 
that some winds blew more persistently of a morning than of an evening, 
while the reverse was the case with others. This brought up the thought 
of land and sea breezes, so the original object was abandoned for a time, 
and the present paper was the result. 
In dealing with this question the wind rose has been divided into 
eight points : north winds include those from the NNW. and NNE. ; 
west winds those from the WNW. and WSW. ; south winds those from 
the SSW. and SSE. ; and east winds those from the ESE. and ENE. 
Pairs of wind directions will be considered separately; the argument 
as to what should take place will first be stated, and thereafter the proof 
or otherwise deducted from actual observations and tabulations applied. 
This is a correct scientific mode of procedure. 
North and South Winds. 
The north is a decided land wdnd and the south a sea one ; of that 
there is no doubt, and we should expect to find more of the former of a 
morning and of the latter of an evening. Such, in fact, is borne out by 
the observations, for the north wind blew at 8 a.m. upon 169 out of 212 
occasions, while the south wind was blowing at 8 p.m. upon 57 out of 87 
occasions. 
