The plug is then to be reinserted and glued in place.

The stand (which it is well to make and fit to the tube before the latter is filled) is shown in Fig. 364. It may be of walnut, mahogany, or cherry, and its general style and finish must depend on the taste of the maker. A shallow groove down its centre receives the tube /, and an oblong cavity at the bottom admits the back of the cistern, while its front may be covered with a hollow ornamental turning g as represented. The scale A, which in this case should be 5 in. long, may be ruled on ivory as already suggested, though an instrument of this description is really deserving of a well-made engraved scale, with a vernier giving readings to the hundredth part of an inch. Such a vernier i is a narrow piece of ivory 1 1/5 in. long, provided with a groove to receive the inner edge of the ivory scale along which it slides next to the tube, a hollow being cut in the wood of the stand behind the scale to admit of its motion. It is divided into 11 equal parts by 10 horizontal lines numbered downward from 1 to 10, each of the divisions measuring therefore 1/10-1/100 in. The 30-in. line of scale is to be placed exactly 30 in. above the centre of the hole e, which marks the level of the mercury in the cistern. It is best to affix the scale to the stand by little brass screws.

A small thermometer k, opposite the barometer scale, adds to the elegance and efficiency of the instrument. A slip of green surface paper should be pasted in the groove behind the tube before the latter is fixed in place. The top of the tube f should be covered by a small turned button l of bone or wood. (A. F. Miller.)

(ft) There is no occasion to wash out barometer tubes at all if new, and old should on no account be used; the only cleaning required, and all damp inside the tube must be avoided most carefully, as it is almost impossible to dry it afterwards, is a clean bunch of cotton on end of a string passed through the tube once or twice, or swab of linen. On no account pass wire of any kind through glass tubes, as if you do the tube will infallibly break in a short time. The longer limb of the siphon need only be 32 in., as the range of barometer never exceeds 31 in. The extra length represents a terribly extra weight of mercury, and as that metal is very dear, a good many pence are saved. The mercury, if bought from a respectable firm, is quite pure enough without being distilled, but on no account must damp be allowed to get to it before being placed in the tube. The best way of filtering the mercury is to take a piece of writing paper and well dry it before a fire, then twist it into a cone with a very fine hole at the bottom exactly like the grocers do up sugar in; a little knack is required for this, but will soon be acquired. Now comes the worst job of all, the filling. The usual method is to shake it in, but it is a most difficult thing to describe.

The proper way is to introduce an inch or two of mercury, very gradually warm the tube, and boil the mercury; allow it to cool, and add another inch or two; warm, boil, and allow to cool; and so repeat till full.

But I have never seen a siphon tube boiled; the air is thoroughly shaken out, until, when the tube is slightly inclined, the metal strikes the top of the tube with that peculiar click which shows all air is expelled.

If amateurs take my advice, and do not boil,, they will save themselves a good deal of trouble and loss, as very likely 3 or 4 tubes will jly during the boiling. Thus there is loss of tube, and, what is more expensive, mercury; and I will promise that if the air is well shaken out, the tube will read within 2/100-3/100 of a standard barometer. Of course these remarks do not apply to pediment or standard instruments.

The wooden cisterns are screwed, and the leather is laid on the bottom, being cut to fit inside the flange. Now fasten the tube in the top, the end of the tube being level with the bottom of cup-like depression in upper part of cistern. When fastened, fill the cistern with mercury up to level with top of cup, part screw on bottom with the leather previously fitted, and you will find it securely held between the two surfaces of the wood. Glue a slip of paper round outside of joint, and it is finished. There need not be any hole inside if done this way, and therefore no glue used that is at all likely to touch the mercury, as if once damp gets in the instrument is ruined.

(10) The vernier should be 1 1/10 in. long, and this should be divided into 10 equal parts, each part would then be equal to 1/10 in. by 1/10 of 1/10 in., that is to say to 1/10 in. by 1/100 in. or 11/100 in.

(11) Camphor Barometer, or " Storm glass." It was discovered by our ancestors that the height of the undissolved camphor in a camphor bottle was different in different states of the weather, and it became quite customary to keep a camphor bottle in sight, in order to anticipate the change in the weather, it being thought that an increase in the height of the camphor indicated approaching rain or wind. The inventive genius of our age could not long allow the instrument to remain in this crude form. The camphor * and alcohol were put into long glass tubes hermetically sealed at the top, and adjusted in a frame with a thermometer attached. The side of the frame next the tube was divided into three divisions. On the bottom one was marked the word " Fair,'* on the middle' one " Change," and on the top one " Storm." A note accompanying the instrument stated that the weather was indicated by the word in the division with which the top of the chemical substance corresponded. It also stated that the direction of the wind was shown by the substance being a little higher on the opposite side from which the wind came.

These were manufactured and sold all over our country under the name of "Storm Glasses," or " Chemical Barometers." One firm, adopted the name of " Signal Service Barometer," and a number of people in possession of these instruments really think they have one of the genuine barometers which the Signal Service use in its predictions.