ON THE METEOROLOGY OF PORT LOUIS. 113 



maximum, however, being considerably greater than the August maximum, 

 and the November minimum considerably less than the June minimum. 



We have seen that the diurnal march of the humidity corresponds with 

 that of the temperature in an inverse sense, the coldest hours being the 

 moistest, and the warmest hours the driest. Such is not the case with the 

 annual march, for the most humid months are the warmest. This seems to 

 arise from the greater length of time that the high temperature prevails in 

 the one case than in the other. Prom August (which is nearly the coldest 

 month) to November, the humidity goes on decreasing with an increasing 

 temperature, the relation between the two elements in their annual march 

 being here analogous to that between them in their diurnal march ; but it 

 would appear that by December the vapour has accumulated so much that, 

 notwithstanding the increasing temperature, the humidity, instead of de- 

 creasing further, begins to increase, and it goes on increasing till February, 

 Owing to the excess of accumulated vapour, time is now required to restore 

 the two elements to their normal relation ; and although the temperature 

 decreases, the humidity does not increase but decreases, the evaporation from 

 the surrounding ocean becoming less active ; and it is not till June that the 

 humidity begins to increase with a decreasing temperature. In August the 

 temperature commences to rise, and then the humidity decreases with the 

 increasing temperature till November or December, when the overpowering 

 effect of evaporation again causes the humidity to increase with the tem- 

 perature. 



Extreme Monthly Range. — Table XIX. gives the highest and lowest 

 humidity and the extreme range for each month. The greatest range for 

 any one month was 38-7 in January 1860, and the least 17"5 in November 

 1866. January, February, March, and August are the months in which 

 the greatest fluctuations occur. 



Secular Variation. — The last two lines in Table XYIII. show that 1860 

 was the most humid year (73-6), and 1866 the driest (66"4) ; and that, upon 

 the whole, the humidity, Hke the vapour-pressure, has been decreasing since 

 1860. The most humid month during the seven years was August 1860 

 (77-8), and the driest November 1866 (57-0), when a severe drought pre- 

 vailed. 



Extreme Annual Range.— Tah\e XX. shows the maximum and minimum 

 relative humidity, the epochs, and range for each year. The mean annual 

 range is 41-5. 



IV. Atmospheric Pressure. 



Diurnal Variation. — Table XXI. exhibits the mean pressure of the atmo- 

 sphere for each of the hours 3| a.m., 9^ a.m., 3| p.m., and 9| p.m. ; and 

 whether we regard the monthly results, or the yearly results, for those 

 hours, we find two maxima and two minima, the maxima occurring at 9| 

 A.M. and p.m., and the minima at 3| a.m. and p.m. From the last hue but 

 one it appears that from 3| a.m. to 9^ a.m. the barometer rises from 30-038 

 to 30-086 inches, which gives a range of -048 inch; from 9^ a.m. to 3| p.m. 

 it falls from 30-086 to 30-015, that is, to the extent of -071 ; from 3^ p.m. 

 to 9| P.M. it rises from 30-015 to 30-085, or to the extent of -070 ; and 

 from 9^ P.M. to 3i a.m. it falls to the extent of -047. 



But in order to know with certainty whether the march is a double progres- 

 sion, and, if so, what are the epochs of the turning-points, we must examine 

 the term-day observations. The results of these for four years are pre- 

 sented in Table XXII. The last two columns exhibit the mean atmospheric 



