114 RE PORT 1 867. 



pressure for each hour of the day, and its deviation from the mean of all 

 the observations. Beginning with 9 a.m., we find that for that hour the 

 mean height of the barometer is 30-090 inches. It then gradually falls to 

 30-017 at 3 P.M., from which hour it ascends till 10 p.m., when it stands at 

 30-086. It again gradually falls to 30-037 at 4 a.m., from which hour it 

 again rises till 9 a.m. We thus see that the diurnal march of the atmo- 

 spheric pressure is a double progression with four turning-points, namely, 

 two maxima at 9 a.m. and 10 p.m., and two minima at 4 a.m. and 3 p.m. 



This diurnal oscillation of the atmospheric pressure at Mauritius, as at 

 other tropical stations, is extremely systematic and regidar. Its amount, 

 and the epochs of its turning-points, vary a little according to the time of 

 year, as may be seen from the Table ; but, except on very rare occasions, as 

 on the 13th of January 1863, when the centre of a revolving storm was pas- 

 sing near the Observatory, it makes its appearance unerringly in aU kinds of 

 weather. Several theories have been framed with the view of explaining it, 

 but none of them has met with entire acceptance. 



Annual Variation. — In the last two columns of Table XXIII. we have 

 the annual march of the atmospheric pressure, and the monthly deviation 

 from the mean for the year. We perceive that the mean pressure for 

 February is 29-843 inches, that from February to August it gradually in- 

 creases to 30-193, and then gradually decreases till February, and that thus 

 the progression is single, having one maximum and one minimum. The 

 annual march of this element, therefore, is in a contrary direction to that of 

 the temperature, the maximum of the one corresponding nearly, but not 

 exactly, with the minimum of the other, and vice versa, the turning-points 

 of the atmospheric pressure occurring later than those of the temperature. 

 From December to April inclusive, the barometer is below its mean for the 

 year (30-056), and during the other months above it, the epochs of the mean 

 being about the 11th of May and the 9th of November. 



Extreme Monthly Range. — Table XXIV. gives the maximum and minimum 

 pressure and the range for each month, with their means. The greatest range 

 in any one month was 0-977 inch in February 1861, and the least "170 inch 

 in December 1860. December, January, February, March, and June are the 

 months in which the greatest fluctuations occur. 



Comparing the mean monthly oscillation of the atmospheric pressure given 

 in this Table M'ith that of the vapour-pressure in Table XIV., we find that, 

 if the oscillations of the vapour-pressure aff'ect the barometer to their full 

 extent, the barometric oscillations depend more upon those of the vapour- 

 pressure than of the diy pressure. 



Secular Variation.^On looking over the last line in Table XXIII., show- 

 ing the annual means, we find that since 1861 the atmospheric pressure has 

 been increasing. The lowest annual mean is 30-032 for that year, and the 

 highest 30-081 for 1864 and 1866, which gives a range of -049 inch, an 

 amount no doubt mainly due to the disturbing effect of hurricanes. Owing 

 to the great prevalence of hurricane weather in February 1861, for ex- 

 ample, the mean for that month (29-665) is less than it would otherwise 

 have been, and consequently that for the year. 



Extreme Annual Range. — Table XXV. gives the greatest and least pres- 

 sure, the epochs, and range for each year. The mean annual range is 0-918 

 inch, while that of the vapour-pressure (Table XV.) is 0-494 inch. 



V. Pressure of the Dry Air. 

 The phenomenon of the double maximum and minimum, exhibited by the 



