316 
On the Construction 
[Nov. 
tlie lower hole, that the barometer was vertical. By moving a little to the right or 
left, the scale was easily aligned with the plummet in one direction, and its vertica- 
lity proved, while in the other they appeared parallel, which in dimensions of nearly 
three feet, allowed of the verticality being confirmed, certainly within half an inch, a 
deviation which would not occasion an error of more than 1000th of an inch. This 
confirmation was only resorted to occasionally when any suspicion of derangement 
occurred, though as it did not occupy a quarter of a minute, there was no reason 
why it should not always be practised. X should add, that it was this susceptibi- 
lity of check at the moment of observation, which induced me to prefer the plum- 
met to a spirit level, which I at one time thought of applying. 
The barometer being now vertical, and the scale adjusted to the surface of the 
mercury, the height of the column was observed and read off. A little time was al- 
lowed it to settle, when it was again read off : the latter reading being recorded. If 
much difference occurred, a third reading was taken : andlastly the temperature ofthe 
mercury in the cistern was observed, being assumed to be the same as that in the 
tube. The tube was now removed from the cistern, the mouth closed with a cork, 
or piece of leather tied over it, and replaced in its case. The. mercury was returned 
into the bottle, which with the cistern were repacked in the box, and’ the case being 
removed from the square tube, the latter Wits also separated by means of its screw, 
when tlie whole became portable again. 
I have given the above description, because I think it may be of use to those 
who are able to procure unmounted tubes. They may in this way convert them 
into portable barometers at a very small expense of means, and in any part of the 
country, however unskilful the workmen ; I have therefore at the hazard of being 
prolix tried to make it as clear as I could. To some I fear the use of such an ap- 
paratus may appear to involve a very tedious and troublesome operation. But I can 
assure them that so far from it, the whole could be performed in much less 
time than they will read the details. I bad, however, a great convenience in 
performing all these little operations by being enabled to place the whole ap- 
paratus on the top of a box, which invariably accompanied me for other purposes, 
and which made the height so convenient as to allow me to adjust every thing with- 
out any constraint of position, one of ttie greatest enemies to expedition, and I will 
add, to accuracy of observation. And the improvementl have suggested at the com- 
mencement of this paper, having occurred to me, and been adopted, (though forwant 
of tubes of sufficient length, I was prevented doing justice to it,) I found the opera- 
tion so shortened, that I could see no objection whatever remaining on this score. 
But it is to be considered, even supposing the reverse were the case, which is the 
just economy, by devoting a little more time to a work to finish it so effectually 
that it shall not require revision, or for the sake of expedition content ourselves with 
observations that are of no real value, and from which no correct result can beobtained? 
An objection is sometimes heard, as being made to accurate methods, that they 
consume time ; but the very reverse is the truth, they save time in the end. At least, 
it is on such a conviction that I found my recommendation of the instrument I have 
described. 
1 lie circumstance of the tube being entirely exposed, and visible in all its length, 
I consider one of the most valuable features of this kind of barometer : at the mo- 
ment of observation it is seen without any trouble, whether or not the barometer is 
m perfect order. So that should an air bubble by any means occasion a break 
in the column, the mischief is at once detected and remedied before a whole series 
of observations has been vitiated, and the time and trouble expended on them lost; 
or still worse, opportunities of observation passed away that cannot be recalled. 
And what is still more important, the error is not only detected, but may be correct- 
ed on the spot, if a few embers can be got for the purpose of re-boiling the mer- 
cury m the tube. A barometer which may not at any time, at the moment of ob- 
sei ration for instance, and without trouble, be scrutinised, and if found defective, 
corrected, cannot be considered otherwise than imperfect. No person who values 
I'^time will consent to he encumbered with such an instrument. 
There is still another objection to the barometer, as ordinarily constructed, which 
floes not apply to this form of the instrument. This is the limit to their scale 
, . . - seen oy juessrs. unoerr ana uo. only to 21. When 1 or 
dered my brass scales from Dollond,! hud them divided down to 17 inches, which . 
imagined, then, wouldbe the limitoftheirtravels upward. However it proved otherwise, 
lor both scales have visited elevations where the mercury sunk 4 inch below this 
