THE CLIMATK OF LOURENl^O MAROUES. 31 3 



Knowledge of the hydrometrie condition of atmosphere is a 

 vakiahle element owing to its depressing and disagreeable in- 

 fluence, especially in high temperatures. In Loureneo Marques, 

 as we have seen, the excessively dry hot zcijid is very un|)leasant, 

 although it is un([uestionably less prejudicial to health than a warm 

 damp wind. The greatest percentage of relative humidit\- cor- 

 responds to the first 5 months of the year, and the smallest is 

 recorded from July to October, which are the months of lowest 

 temperature and very scarce rains. 



The excess of humidit\- from March to May contri]jule> to 

 make these months worse than what might be expected if we 

 only took into consideration their comparatively low temperature. 



If we compare the diagrams of monthly temperatures with 

 those of relative humidity we shall see that they follow ditterent 

 directions, but there exists no ]>erfect relation Ijetween them, as 

 there is a certain retardation in the march of the humidity curve. 

 Thus in .\pril, when the decrease of temperature is already 

 noticeable, humidity attains its maximum, and then it begins to 

 decrease until September, when it reaches its minimum, while the 

 temperature rises from July on. 



In relation to the rainy season relative humidity attains its 

 maximum at the close of the season, and the minimum is reached 

 in September just after the amount of monthly rainfall begins to 

 increase. 



The fact that sea winds prevail in the hot season should also 

 have its influence on the ranges of relative humidity, as according 

 to observation the dampest winds are those of NE to SSW 

 throtigh the E. 



In theory, relative humidity, /. Ijeing the ratio between vapour 

 tension f (absolute humidity) at a certain moment, and Aapour 

 tension F at saturation, should decrease according to tempera- 

 ture, as the numerator does not increase much, while the de- 

 nominator becomes greater. 



We have already seen how a retardation occurs in the annual 

 variation in relation to the march of temperature ; let us now see 

 the latter's influence on the diurnal variation. 



The maximum relative humidity precedes sunrise, and coin- 

 cides with the thermometrical minimum ; the minimum generally 

 occurs at about noon, before the maximum temperature, and the 

 decrease follows regularly the rise of temperature. The increase 

 in the percentage of humidity is more rapid from the hour of the 

 minimum until the XXI ; afterwards it slowly increases and 

 saturation, or near approaches thereto, are recorded. 



The absolute minimum of relative humidity corresponds to 

 hut zcind days, and it goes down as far as 8 per cent. 



The mean annual variation is ^7,, and siniilarly to the varia- 

 tion of temperature it is greater in the coolest months, the same 

 occurring with the difl'erence between the absolute maximum and 

 minimum. Sometimes a raj^d increase in relative humidity 



