1829 .] 
Of a Portable Barometer . 
317 
point. At each place an observation was made, and the record permanently pre- 
served, by making- a mark with the point of a penknife on the scale opposite the 
summit of the column. At leisure the quantity was accurately measured, by which 
it fell below the division marked 17. In a similar case, where a barometer of the 
more usual consctruction had been employed, the enterprising traveller was deprived 
of the satisfaction of knowing to what elevation he had ascended: the mercury 
sunk down and disappeared within the closed part of the case. 
As before observed, 1 always thought that the instrument would he something more 
convenient in use, if the tube were fashioned into a syphon shape; and the little ex- 
perience I have had of such tubes, strengthens the opinion I had previously formed 
in their favor. The straining of the mercury, the tilling and emptying of the cistern, 
are all got rid of by this means ; as also the trouble of carrying so many additional 
pieces. Nor do I think the difference in charge could be very great, or if it were, 
the purchaser could order straight tubes, and himself give to them the syphon form. 
This was the course I pursued, but the tubes not having been ordered with refer- 
ence to this operation, were deficient in length, so that 1 was not able to give the 
plan a fair trial. 
An objection is sometimes made to the syphon barometer, that the scale being 
fixed, the motion of the mercury is reduced to one half of what it would otherwise 
be; while if the scale be moveable, two observations are required instead of one, and 
consequently there are two chances of error instead of one. To the first objection 
I answer, that the scale ought not to be fixed ; and to the second, that it is so far 
true that two observations must be made, I will not say instead of one, for l assert 
that it is the impossibility of making the second observation that, renders so many 
barometers useless, except us mere playthings. This remark does not apply to 
Troughton’s, Dollond’s, or Berge’s. 
Under the idea that some of your readers might wish to order from England a 
barometer such as I would recommend, I have prepared a drawing of one (1*1. 
IV.) and shall give a description of such part of it as appears to require it. 
Fig. 1 represents" the instrument without its case. The tube is 42 inches long, and 
\ of an inch internal diameter. It is bent into a syphon shape, so as to leave one 
leg 32| inches in length, the other 8£. A brass scale 31 inches in length, divided 
down to 15 inches, to tenths of an inch, is attached to it in the manner shown in 
the plate, with the addition of a screw ring set on at the height of 14 inches from 
the heel. The apparatus shown in the plate, as attached to the heel, is a screw for 
slow motion ; the modus operandi is obvious. When the heel of the scale is brought 
near the surface of the mercury, in the shorter leg, the clamping nut a is fastened, 
when the screw b will be found to act on the scale, and adjust it to the greatest 
nicety. To the heel of the scale should he attached a thin plate ot brass, bent so 
as to embrace the tube, from the alignment of which with the surface of the 
mercury the most correct adjustment may be made. A small piece is cut 
out of the heel of the scale for the purpose of hitching on the shoulder of the 
ring ; but it would I think be preferable to make the latter, with a projecting 
piece to take under the heel of the scale, which should be left even. Another in- 
advertence in the elevation I may notice. It is the position of the slow moving 
screw c, which to allow the instrument to pack conveniently, ought to be in front. 
There will be no difficulty in making these slight alterations, lhey are corrected in 
the section fig. 2. where the same letters refer to the same parts. 
The scale. I would have only divided to lOths of an inch, the vernier subdividing 
to lOOths, it will be so much cheaper, and I am convinced that estimation assisted 
by a magnifier will subdivide thelOOtbs of an inch as accurately as is required, fully 
as much so as any barometrical scide 1 have seen. The reading is made by this 
means more simple and direct, and in observations which like these are sometimes 
required to be taken in a hurry, that is a point of some consequence. 
On the continent, the stop-cock which used to he fixed at the bend, has given way 
lately to an improvement suggested by Gay Lussac, of forming the bend into a ca- 
pillary tube, thereby preventing those shocks to the mercury which occasion so of- 
ten the intrusion of air bubbles, and even of arresting the progress ot the latter in 
a large tube ; this is a decided and a great improvement, but such tubes as I used 
were already capillary, and scarcely required any reduction in their caliber. Whether 
such a modification would be advisable in a tube off inch diameter! can scarcely 
give an opinion. If Dollond or any of the superior makers be employed, it might 
be perhaps left to their judgment. 
As to the manner of setting up such a barometer, I would propose the following 
improvement of the apparatus described above. 
