3J'' THE CLIMATE OF LOURENCO MARQUES. 



correspond to a dehned increase in the amount of rain. In 

 short, rainfall in Lourenco Marques, followino- the type of 

 tropical rainfalls to the season, is closely connected with the pass- 

 mg of depressions, and approaches the rainfall known as cyclonic 

 rainfall. 



When in the isol)ars the centre of hioh pressure is estab- 

 lished at Lourenco Marques, small showers are sometimes re- 

 corded with changing winds, which is in disagreement with the 

 theory of " convectional " rainfalls, because, as is known, 

 descending air currents correspond to the centres of high ores- 

 sure, and such currents are the least likely to promote rain. 



In Lourenco Marriues rainfalls are very fre(|uent from TV to 

 IX and after XVIII : during the remaining hours it rains more 

 at night than during the day, and it rains the least from XV 

 to XVII. that is to say, when the velocity of the wind is greater. 



During four years — 1909 to 1912 — on 20 different occasions, 

 rainfalls have been recorded which registered over 30 millimetres 

 in 24 hours. Of these the highest figures recorded by the rain 

 gauge are as follows : 



In 1909 — January 70'""' {ij^y" ) in 1 hour 15 minutes. 

 November 30""" (i.iSi") in 15 minutes. 

 December 50""" {1.969") in 15 minutes. 

 December 96""" (3.781") in 4 hours. 



In T910 — February 18T"'" (7.128") in 12 hours. 

 October 43'"'" ( r.6(j3") in 40 minutes. 



In 1911 — October 50'"'" {].</«/') in 1 hoin-. 



Together with the percentage of sunless hours and cloudi- 

 ness, I give a diagram of rainfall which I consider interesting on 

 account of the general impression it affords as to the connection 

 of these three elements, and of the general condition of the atmos- 

 phere. 



I also introduce a tliagram showing the average monthly 

 number of rainy days and thunderstorms. 



Evaporation. 



A " Piche " evaporimeter is employed at the Observatory to 

 measure the evaporation, the- instrument being read at IX. As 

 is known, the study of evaporation is not well defined, it being 

 impossible to compare the observations made with the same in- 

 strument in two different localities, this being principally due to 

 the influence of the temperature on the water. It is admitted 

 that evaporation increases with the temperature of the water 

 and the velocity of the wind, and decreases with humidity and 

 atmospheric pressure. This is what actually takes place in 



Lourenqo Marques, wdiere the influence of the temperature and 

 the velocity of the wind are quite manifest, both on variations 

 from day to day and from month to month. This latter factor 

 is, however, the prevailing one, as evaporation is greater during 

 the hot season, in the month of greater wind velocit}-. In the 



