the Trigonometrical Station on Rumbles Moor, Yorkshire. 133 
Great Almias Cliff. 
1821. h. m. h. m. Therm. Wind. Bar. 
Aug. 15, 15 obs. from 12 30 to 16 30—62to64 W. 
Jack Hill. 
Aug.17, 8 .. .. 16 Oto 17 0—57 S.W. 
Symon Seat. 
Aug. 29, 8 .. «. 15 20.,16 40—52..53 S.E. 28-417 
Sept. 3, 9 ee ee 14 0.. 15 30—56.5.07 S.W. 
: Great Whernside. 
Sept. 1, 9 .. «. 1210..18 45—52..56 N.W. 
Pendle Hill. 
Sept. 24,12 .. .. 1050..14 20—48..51 W.N.W.28-050 
1822. Alfred Castle. 
Jan. 26, 17... «2 11 0..13 15—40..41 W.N.W. 
The first fact elicited by these numerous observations was the 
existence of a species of diurnal variation of refraction not exceed- 
ing 60" to 70” within the limits of the survey, and dependent on 
the locale of the station, the time of the day, and in some degree 
_ (if it may be so termed) on the constitution of the day itself. It 
was first noticed at the observatory, Jan. 25, 1821, and might 
be said to be at its maximum in March, at which period it ex- 
ceeded 60”. At Rumbles Moor it was still more marked up to 
the 3d of June, when it totally ceased. No observations ulteriot 
to those just mentioned, were made at the observatory in 1821 ; 
but such angles as have been taken in January last, give no indi- 
cations of its return. It is also worthy of note (and a comforta- 
ble discovery it is for the surveyor) that the mean of the diurnal 
extremes differs but very slightly from the constant angle; gene- 
rally speaking, the variation is greatest when the mornings are 
frosty, and the sun acquires great power during the middle of 
the day. This led me to suspect, that in spite of the sector 
being reversed after every observation, the variation might be 
wholly attributed to the change of temperature as affecting the 
instrument. Observations made on frosty mornings with the 
sector at 30°, and afterwards heated to 60°, proved by their 
near agreement that the cause of the variation could not be 
looked for there. Asa further and irrefutable confirmation, the 
refraction was constant at Beamsley Rock, May 2, (on’ which 
day the thermometer had an extensive range,) although the va- 
riation continued to be observable at Rumbles Moor, It is to 
be 
