312 LIEUT.-COLONEL SYKES’S DISCUSSION OF METEOROLOGICAL 
for an hour to an hour and a half; nevertheless the monthly means of the same 
tide for successive years are almost identical ; take for instance the rising afternoon 
tide from 3 — 4 p.m. to 10 or 11 p.m. in the month of January, and it is seen that the 
means in the successive four years were *078, *080, *078 and '079, so that the daily 
irregularities were merged in the monthly means. Similarly, the rising tide from 
3 >' 41 “ — 4 ^ 41“a.m. to O’* 41“ — 10’* 41“ a.m. in the month of February was '096, '095, 
•095 and '100 in successive years. The falling night-tide between p.m. and a.m. does 
not exhibit quite the same regular features ; for instance, in the month of March we 
have ’064, '058, '055 and *071. November also has '063, *070, '059 and '075 ; but the 
annual means of each tide, as in the great diurnal tide, remarkably approximate to 
each other in successive years, as is shown by the Table, which is worked out from 
the Madras hourly observations. Another feature exhibited by the Table is the alter- 
ation in the range of the monthly means of the diurnal movement of the same tide 
in different months of the year, the increment and decrement alternating at different 
seasons of the year with the increment and decrement of another of the tides; for 
instance, in the month of January throughout the four years the rising tide from 3’* 41“ 
— 4’* 41“ P.M. to 9’* 41“ — 10’* 41“ P.M. is less than the rising tide from 3’* 41“ — 4’* 41“ 
A.M. to 9’* 41“ — 10’* 41“ A.M. ; but the reverse of this takes place as the year advances, 
and in April the p.m. rising tide is considerably greater than the a.m. rising tide, and 
this continues until November, when the previous relations return. The same remarks 
do not apply to the two falling tides, as the falling tide by day is invariably greater 
(more than double) than that of the falling tide by night. 
In my paper on the Atmospheric Tides of the Deccan, I gave a Table (No. 3) of 
the anomalies in the period of the ebb and flow of the different tides in 1830, and to 
enable me to do this I had observed the barometer every five minutes about the period 
of the expected turn of the tide. The anomalies were numerous. In the Madras 
hourly observations the same anomalies are observable, as the intervals of maxima 
and minima vary from five hours to eight hours, although the absolute time of the 
turn of the tide within the hour is not shown. Dr. Buist’s observations in Bombay 
for 1843 and 1844 are also hourly, and exhibit the same occasional irregularities in 
the time of the tides turning as at Poona and Madras. The recent observations at 
Calcutta do not afford the means of detecting this no doubt existing fact ; for ob- 
servations at Calcutta are not recorded by night at all, and only every two hours 
during the day. A bare inspection of and reliance upon the Calcutta monthly tables 
therefore would justify the assertion that the daily maxima and minima in the oscil- 
lations of the barometer occurred at fixed hours, which, although generally true, is not 
absolutely so. The observations at Dodabetta were unfortunately not made hourly, 
with the exception of those for twenty-four hours between the twenty-first and twenty- 
second days of each month ; the means of comparison therefore of the diurnal hourly 
oscillations with those of Madras and Bombay, are limited to those twenty-four hours 
in each month ; but although so limited, features of interest are exhibited. It has 
hitherto I believe been supposed that the two ascending and two descending diurnal 
