590 MR. HUDSON’S HOURLY OBSERVATIONS ON THE BAROMETER 
from 6 a.m. to midnight ; the mean variation at 8 a.m. is referred to a general 
mean derived from all observations made from that hour till midnight ; and 
the variations at each of the subsequent hours are referred to a mean of all 
observations made from 9 a.m. till midnight*. It points out, — 
7. That the greatest of all the mean variations is nearly "016 inch in 
amount, and occurs in the afternoon minimum height of the baro- 
meter at four o’clock ; the next *012 inch, and found in the forenoon 
maximum at ten o’clock; after this the one of nearly *011 inch in 
the evening maximum at eleven o’clock ; and finally that of -005 
inch occurring in the morning minimum at half-past four. 
8. That the general relation between the barometrical changes and the 
variations of temperature, appears to be direct during the morning 
hours, and inverse during those of the day and evening. 
9. The singular fact, that while a period of fifteen days gives a mean day 
generally distinguished by its relative variations at noon and mid- 
night, a period of one month, or a complete lunation, not only gives 
a gradual succession of variations, but, in all these observations, a 
result almost identical in character and amount with the combined 
result of the whole. 
Among the investigations in which I am at present engaged, are those 
relating to the following inquiries: 1. To ascertain whether the mercurial 
vapour in the vacuum chamber of the barometer, sensibly influences the 
height of the column at the ordinary variations of the temperature of the 
atmosphere. 2. Whether the Tables for the reduction of the temperature of 
the mercury to zero are practically accurate. 3. A full investigation into 
the influence which the diameter of the tube exercises over the fluid column. 
■4. The relation between the mean daily variation of the magnetic needle and 
that of the barometer ; and whether the former would be found to exhibit the 
same dependence upon changes of temperature as the present observations 
have shown the latter to have. 5. The connexion between the mean baro- 
metric height and the amount of the variations referred to it, and the influence 
' I find that a mean derived from all the observations of the twenty-four hours, compared with one 
derived from all those of the sixteen hours, from 9 a.m. to 12 f.m., of the same period of observation, 
differs from it only by - 001 of an inch. 
