Page:Madras Journal of Literature and Science, series 1, volume 6 (1837).djvu/57

1837.] quotient arising from the division of the number expressing the tension of vapour at the mean temperature of the dew-point, by that expressing the similar tension at the mean temperature of the air. The sixth or last column, under the head hygrometer, affords, in connexion with table II. data for calculating the amount of evaporation.

With respect to the quantity of rain recorded in the table, it is less than what usually falls during the corresponding period. The average annual fall of rain at Moulmein is above two hundred inches. Formula have been given for finding the mean fall of rain at any latitude, but it cannot be doubted that besides the distance of a station from the equator, its distance from the coast, the direction of the line of coast, the relative positions of the sea and land with respect to that line, and probably many other circumstances unite in influencing the amount of rain.

The times for recording the barometer were selected as furnishing its two maximum and one of its minimum heights. It did not appear however, that there would have been any sensible difference between the mean of the observations at 9½ and 10½  A comparison of the consecutive monthly differences of the barometer at the three daily periods of observation shews that, at each of these, the column of mercury decreased without intermission from January to June inclusive; in July, it again increased a little. The fall of the barometer may possibly have been caused by the shifting of the prevailing winds from the N. E. and N. W. to the W., S. W. and S. as indicated by table II., but there is nothing either in this, or the first table, to account for the increase of pressure in July.

With regard to the daily oscillations of the barometer, their magnitude is undoubtedly affected by local causes, but these latter ought perhaps only to be regarded as accidental disturbances of those general and constant causes on which the oscillations themselves depend. It appears, from the table, that both of the oscillations recorded increased from January until the end of March, and then diminished: in other words that though, as has been observed, the mean height of the barometer decreased, from month to month, at each of the three periods of the day, its decrease at 3½ was more rapid, until the end of March, than at 9½  and 10½  The oscillations at a given latitude are supposed by M. Bouvard to vary in magnitude directly with the temperature, on the centigrade scale, of the period during which they occur. Is it not possible that they may also vary with the tension of the atmospheric vapour? It will be observed, on reference to the table, that this tension was greater at 9½ and 9½  than at 3½  from February to May inclusive, while on the contrary it was less during January, June and July.