BAROMETER 327 eter. This stand folds up as seen in fig. 4, and serves then as an envelope to protect the in- strument. Mr. Greene constructed also, at the suggestion of Prof. Henry, a sulphuric acid barometer for the Smithsonian institution. As this liquid is much heavier than water, the tube was only about 18 ft. long ; but experi- ence proved it to be behind the mercurial barometer in its indications, and its use was abandoned. The siphon ba- rometer of Gay-Lussac, im- proved by Bunten of Paris, is a very portable and conve- nient form for the use of the scientific traveller. It is rep- resented in fig. 5. The name siphon is applied to barome- ters of which the lower end of the tube is turned tip to form a short arm, which con- stitutes the cistern, and may be left open for the air to press directly upon the mercury. A capillary opening in this short arm, which is otherwise tight, answers the same purpose as if the whole were open. The surface of the mercury in the lower arm cor- responds to the zero point in the cistern ba- rometer; and as this fluctuates as Fiy.5. we ll a8 that of the longer limb, it is necessary to use a vernier at each ex- tremity of the column, and take two readings in order to determine the height of the column. As the two limbs are made of precisely the same diameter, the reading of one and doub- ling this gives a correct result. In Gay- Lussac's barometer, the tube at each extremity is of the usual diameter, but in the elbow, and along the lower part of the long limb, it is drawn down to a very small bore. The instrument is thus made to occupy very little space, so that the glass is enclosed in a brass cylinder of the size of an ordinary cane. An open slit at each end of the brass tube affords an opportunity of reading the verniers, the indexes of which traverse up and down these openings by means of toothed wheels which run in a rack made upon the edge of the brass. The improvement introduced by Bunten is in dividing the long limb into two parts, the upper one of which is drawn down at its lower end to a small opening and in- serted into the lower portion, to which it is attached, making again one tube. (See fig. 6.) The object of this conical projection of the upper into the lower part is to form a chamber or trap to catch any air which may be acci- dentally introduced through the short branch, and thus intercept its passage to the vacuum, where by its elasticity it would counterbal- ance to some extent the pressure of the external air. When the ba- Tff.a-.ll. rometer is inverted, the air lodged in the air trap escapes through the short branch by which it entered. A barometer in common use is pro- vided with an index which turns around upon a dial, and points to figures which indicate the height of the mercury, as also to words de- scriptive of the state of the weather, as " Cloudy," "Fair," "Rainy," &c. The index is made to move by means of a string, which passes around its axle, and has at each end a weight attached, the larger one resting upon the surface of the mercury in the shorter limb of a siphon barometer. (See fig. 7.) This is open to the objection that the reading of one limb gives but half the ac- tual effect ; but as the length of the index is several times greater than the radius of the pulley up- on its axis, this objection is really more than counter- balanced. Still, little con- fidence is placed in its ac- curacy in marking the true variations of the column, there being so much fric- tion that slight changes do not affect it at all. The words "Fair," "Variable," "Rain," "Storm," &c., found on the barometer scales, convey an erroneous impression about this in- strument to the uninstruct- ed ; for the barometer does not designate by the abso- lute height of the mercury, but by its rising or falling, the kind of weather we may expect, and this change is not indicated by the index. In filling a tube with mercury, particular care is required that the mercury be free from mixtures of other metals. It is intro- duced into the tube in small quantities at a time, and boiled as each portion is added, the heat being applied to that part of the tube containing the mercury last introduced. By boiling the mercury in the tube in vacua, the air and moisture are most effectually ex- pelled. On inverting the tube when prop- erly filled, its lower end being kept in a basin of mercury, the column sinks to the proper level to counterbalance the atmospheric pres- sure. When the operation has been suc- cessfully completed, the column of mercury presents a bright undimmed appearance, and emits flashes of electrical light in the vacuum above, on the column being made to oscillate up and down in the dark ; and a perfect vacuum is indicated by the clicking sound of the mer-
