HOURLY OBSERVATIONS OF AIR TEMPERATURE AND PRESSURE. 
637 
ditions are also shown to exist by the two Tables before referred to, the sign of 
being in all cases negative at the equinox. 
In connection with this view of the temperature components it may be noticed that 
if instead of reckoning the epochs of maxima as the phase nearest to midnight, that 
nearest noon were adopted, it would be found that the usual time of maximum phase 
of all the components is not far removed from noon, affording strong evidence that 
all these phases are closely dependent on the passage of the Sun over the meridiau. 
For Greenwich the following would be the results :— 
December. 
March. 
June. 
1st component .... 
0 
22*^ 
C 
215 
O 
210 
2nd ., .... 
200 
198 
181 
3rd „ .... 
194 
, , 
(193) 
4tli „ .... 
190 
(193) 
196 
In case of the third and fourth components the figui’es enclosed in brackets ( ) are 
epochs of minimum. 
In conclusion, I would observe that in the coui’se of the preparation of the data on 
which this discussion has been based, I have computed the harmonic components of 
the temperature curves of many other places besides those here specially noticed, and 
that the results obtained from them entirely confirm those now set forth. I am not, 
however, in a position at the present time to present any satisfactory details of these 
investigations, involving as they do very tedious computations, which I have not yet 
been able to carry out in a systematic or complete shape. 
The importance of the subject in connection with the study of the diurnal variations 
of atmospheric pressure cannot be overstated, and in the second part of this memoir, 
which will be devoted to that question, I shall have occasion to revert to it. Mean¬ 
while I would invite the co-operation of persons interested in this branch of physics 
in the collection of data relating to temperature as complete as those now available 
in the case of atmospheric pressure. 
