320 THE CXIMATE OF LOUREXqO MARQUES. 



element worth considering. On the other hand vegetation, as 

 well as evaporation, alters the temperature of the soil. 



It is also worth observing that there seems to exist some 

 relation between the temperature of the soil and the existence 

 and development of certain diseases, such as diarrhcjea and typhoid 

 fever. 



The reading of the thermometers placed at 0.5,1,2,3 

 metres depth is taken at IX. I attach hereto a table (No. 

 XVI 11) of the averages during three years. Naturally there' 

 exists some agreement with the three laws of the temperature of 

 the earth, i.e.. the law of amplitude, that of retardation during 

 the period of the extremes, and finally that of reduction of ampli- 

 tude, which varies with the period luider consideration and with 

 the depth. 



Th u n derstorm s. 



Thunderstorms are most frequent in December, then in 

 (Jctober, November, and May, and least frequent in June, July, 

 and August, this frequency coinciding with atmospheric distur- 

 bances. We have already mentioned the fact that the wind 

 veering to the South when depressions occur, it is often accom- 

 panied by thunderstorms. 



\\'hen the lowest pressures become established towards W 

 and S\\ . and the barograph begins to trace a sinuous line, thunder- 

 storms are common, with showers from W. Violent thunder- 

 storms, which sometimes last for hours, are accompanied l.)y f resli 

 winds of variable directions. They are more common at night- 

 fall and the last from sunrise to XIII. As one sees, most 

 thunderstorms in Lourengo Marques may be classified as depres- 

 sion thnnderstonns rather than heat thunderstorms. 



Lightning without thunder is often recorded one or t\\'o days 

 ])revious to the occurrence of depressions, and their trerjuency 

 follows the same march as the thunderstorms. 



Other Phenomena. 



I give in Table No. XIX various meteorological data, some 

 of which have already been considered. 



Fog is more frequent during the cool season, especially in 

 the morning. Hail is seldom recorded, and that which fell in 

 1912 in the lower part of the town is worthy of notice on account 

 of its abundance. 



Lunar coronw are often observed ; halos, which are not 

 frefjuent, generally mark the approach of rainfall. 



General Remarks. 

 I have endeavoured without excessively enlarging on this 

 paper, to introduce the most important data for the study of a 

 climate, and have followed the indications of Professors J._ Hann 

 and Cleveland Abbe, in harmony with the observations in my 

 possession. 



