ON THE HOURLY OBSERVATIONS OF THE THERMOMETER. 203 
Form of the four different branches observable in the mean 
annual daily curve . 
The preceding investigations furnish us with the following 
elements of the mean annual curve of the 24 hours. 
Mean temperature of the 24 hours by Table VI.52*90 
Mean minimum.49 
Mean maximum .58 
Difference between max. and min. 9 
Difference between max. and mean temp. 6*9 + 
Difference between min. and mean temp. 3*9 — 
h m 
Time of minimum temperature. 5 0 a.m. 
Time of morning mean temperature. 8 9 p.m. 
Interval between min. and morning mean. 3 9 
Time of maximum temperature. 1 0 p.m. 
Intervalbetweenthe morning mean and following max. 4 51 
Time of evening mean temperature. 7 0 
Interval between the max. and evening mean. 6 0 
Interval between the morning and evening mean .... 1051 
Interval between the evening mean and following min. .10 0 
Interval between the min. and following max.8 0 
Interval between the max. and following min.16 0 
The mean annual curve of the 24 hours, beginning and ending 
with the point of minimum temperature is projected in Plate X. 
It may be divided into the four following branches, C C being 
the line of mean temperature: 
1. The morning branch AB ascending from the minimum to 
the line of mean temperature, and including an interval of 3 h 9 m ; 
2. The noon branch B D ascending from the mean tempera¬ 
ture line to the maximum, and including an interval of 4 h 51 m ; 
3. The afternoon branch DE descending from the maximum 
to the line of mean temperature, and including an interval of 6 h j 
4. The night branch E A descending from the line of mean 
temperature to the minimum, and including an interval of 10 h . 
On examining these four branches of the mean annual curve of 
the 24 hours at Leith, we may observe in them a somewhat 
close approximation to the parabola, on which account I 
have been led to calculate the mean annual hourly tempera¬ 
ture for Plymouth, on the supposition that they may be 
represented by the abscissae of parabolas. Taking C C in 
Plate X. as the line of mean temperature, then CA, GD, and 
their parallels may be considered as the abscissae of the respec¬ 
tive curvilinear branches AB, BD, &c, above mentioned. The 
