578 MR. HUDSON’S HOURLY OBSERVATIONS ON THE BAROMETER. 
tube has an internal diameter of 0’53 inch, and the neutral point of the in¬ 
strument is 30576 inches, at 54°. 
The Water Barometer forms the subject of a paper by Mr. Daniell, printed 
in the present volume of the Transactions, and containing- a full statement 
of its peculiarities and the mode of its construction. 
The Mountain Barometer is the property of Mr. Daniell, and is considered 
by him as an almost perfect instrument. It has a tube of O’15 inch and a 
cistern of T2 inch internal diameter, with a brass scale extending to the surface 
of the mercury in the cistern ; and is the first barometer to which Mr. Daniell 
applied the platina guard for preventing the insinuation of air into the vacuum 
chamber of the instrument. Its neutral point is 30080, at 65°. 
The regularity with which the barometer, in tropical climates, proceeds in 
its periodical rise and fall from day to day with almost uninterrupted progres¬ 
sion, has long been observed by our travellers and philosophers. This perio¬ 
dical oscillation, as the parallel of observation becomes more remote from the 
equator, gradually ceases to be obvious in the observations of a single day; 
and in its place we have the violent and irregular movements of the mer¬ 
curial column, so well known in our own and other extra-tropical climates, 
and in which the effect of no constant law is apparent. By classing, however, 
the observations made at the same hours on several successive days, and de¬ 
riving from their union the hours of one mean day, it has been found that 
these accidental variations destroy or neutralize each other, and allow the con¬ 
stant, or equatorial, oscillation to become appreciable and subject to investi¬ 
gation *. The results now presented to the Society consist of eight such mean 
days, each of them derived from observations made on fifteen days, a period 
I have adopted as the standard, and which appears to be amply extensive 
for clearing the result from the interference of the accidental variations. In 
forming each mean day, all the observations made at a given hour, on succes¬ 
sive days, have been collected together, their sum taken, and a mean re¬ 
sult for the given hour obtained by dividing that sum by fifteen, the number 
of the observations. A mean hourly result for the temperature has been 
obtained in the same manner. Having thus derived a mean quantity for each 
* The clear and striking statement of these phenomena, given by Sir John Hehschel in his Preli¬ 
minary Discourse, (§ 228.) suggested the original idea of the present observations. 
