The straight part of the tube, exclusive of the bulb, is 32 inches; inclusive of the bulb, 34; the recurved end is bent twice at right angles, so as to project from the tube 3 inches, and rise parallel thereto 7.5 inches. The tube is attached to a mahogany support, the spheroidal bulb being upwards; and the quantity of mercury is so adjusted in the tube that at mean pressures the upper level is nearly coincident with the greatest diameter of the spheroid, and the lower is near the middle of the shorter leg.

There is a circle of brass, divided into 1000 parts, fixed to the front of a light copper drum or case, having a glass front and back, the centre of which circle is placed just over the orifice of the glass tube: a small frame of brass is fixed to the circle behind, so as to carry a light horizontal axis bearing two small pulleys. The extremities of this axis are turned to extremely fine pivots, and are set in small jewels: the front one projects forward so as to carry a light index of straw, which is sustained on a small brass ring, placed by means of a socket on the extremity of the axis, in the manner of the hand of a watch. The two small pulleys above mentioned carry, by means of fine untwisted silk threads, two small cones of glass or wood, one of which rests on the surface of the mercury in the recurved tube, the other hangs freely on the outside of the tube. These cones are nearly equal in weight, that resting on the mercury being rather the heavier of the two. This slight difference of weight, setting aside the inertia and friction of the axis, is the amount of the resistance which the rising or falling of the mercury in the tube has to contend with; and this is all extremely little, so little that the index moves by the unequal action of the wind during a light gale, and is put into a state of oscillation of some considerable duration by the mere opening of the door of a room. These pulleys measure very nearly one inch in circumference, so that if the mercury moves an inch the index is carried once round the brass circle, and hence one division thereon corresponds to 10th of an inch, a correction being made on the pulley according to the relative capacity of the tube and the bulb.

1

1000

The index is made in three parts, of light straws, a centre piece and two extreme pieces inserted into it: one extremity is cut after the manner of a pen to a somewhat short and very fine point, which is turned edgeways. The whole is carefully equipoised by a short piece of straw sliding on one of the extreme pieces, so that when attached to the axis it takes indifferently any position in the circle, and, consequently, follows exactly the movement of the mercury.

A varnished paper is pasted on the front of the tube, marked 27, 28, 29, 30, &c., to denote the height of the mercurial column in inches; these measures being taken with care from the surface of mercury in the bulb to that in the tube. The index is set as nearly as possible when the altitude corresponds to fixed divisions of the scale or measure.

There is a thermometer close to the mercurial column, the bulb of which is placed in a small cistern of mercury, to indicate the temperature, and a hygrometer to measure the change which may be supposed to happen in the silk line, to which the cone resting on the mercury is attached; but the author has found that by employing fine unspun silk the changes are quite unimportant.

The quantity of mercury in the instrument is about 15 lbs. In order to fill the tube clear of air, the following process was adopted as a substitute for boiling. A small iron cap, polished, was first cemented air-tight upon the end of the tube, and into this was screwed an iron stopcock: a long glass tube was then cemented to the stopcock, furnished with iron caps, &c., so that by reverting the instrument and steadying the tubes by cross-bars of wood, tied with silk-ribbon band, the whole may be screwed into the plate of a good air-pump.

The air being withdrawn as completely as a good air-pump will effect, the cock is closed, the whole is detached from the pump, the long tube removed, and the barometer tube transferred to a cistern of mercury, under the surface of which the stopcock is fairly immersed, whilst the tube is inclined as much as possible. The operator, being placed in a convenient position, supports the ball of the tube in his hand, and turns the cock gradually, so as to allow the mercury to be pressed up in an extremely small stream into the tube, and to flow down without violence into the ball. During this process the ball is gently moved about with an easy circular motion, which allows of the more speedy union of the mercury and displacement of the air. An assistant should be ready to close the cock occasionally, for the purpose of examining the state of the mercurial mass within the tube.

In this way the barometer tube may be filled with great nicety, so as to show a most resplendent surface, equal in appearance to that produced by boiling even under a powerful magnifying glass.

When the tube is complete to the point required, the stopcock is again closed, the whole is reverted, and the tube is placed in its intended place; the cock is then opened by degrees, and the mercury will gradually descend to the level of

the atmospheric pressure. The iron stopcock and cap may now be removed, by cautious application of first a warm and then a hot iron rod to the cement.

The mercury intended for the purpose of the barometer should be first distilled, and then well agitated, about an ounce or less at a time, in phials capable of holding one ounce and a half. Previously to introducing the mercury into the tube it should be well boiled in a crucible, of porcelain or Wedgwood ware, and should be used just before getting cold, at a temperature of 90° or 100°, the tube of the barometer being also made a little warm by careful exposure to a charcoal fire.

The process now described is believed by the author to be, when carefully performed, in every respect equal to that of boiling. The wheel-barometer made in this way has been compared with other instruments with boiled tubes and of undoubted excellence, amongst others with a fine mountain barometer on Gay-Lussac and Renard's principle, which had been compared with the standard of the Royal Society in Somerset House, and with that in the observatory at Paris. The differences from this instrument, when both were placed in the same room, were very minute. It is found to be more sensible than a very finely boiled tube, carefully prepared by that eminent maker Mr. Cox, of Plymouth, with a scale and vernier divided to 5th of an inch, set up in an adjoining room.

On a new Method of Constructing a Portable Barometer. By JOHN NEWMAN, Mathematical Instrument Maker.

The object of this construction is to make barometers portable without the use of a leather bag, which has always been a defective part of the instrument.

The method adopted is to have a cylindrical cistern of iron in two parts, rather longer than usual, the upper part, or chamber, or that to which the cap is fastened, which connects to the tube, being about three times the length of the lower part, of the same diameter, moving round upon a pin, and secured by a screw and collar. The two chambers thus formed communicate internally in one situation by means of holes in the divisions, through which the mercury flows upon inverting the instrument. The vacant space, or that intended to receive the mercury from the tube when the barometer falls, is, when the instrument is in use, in the upper part of the upper cistern, the lower one being full. Upon inverting the instrument, the mercury flows from the latter into the former, which 1833.

2 E

becoming filled, is by a quarter turn of the one now uppermost cut off from communication with it, and the instrument is rendered portable with the end of the tube dipping into a cistern of mercury, which is perfectly secure.

By this method Mr. Newman is enabled to make portable mountain barometers with very large tubes, for sufficient room can be left in the cistern to receive the mercury which flows from the tube into the cistern in high situations, notwithstanding the increased diameter of the tube. Barometers, therefore, can be made and transported, which when put up may be depended upon as standard instruments with perfect security. On an Instrument for measuring the total heating Effect of the Sun's Rays for a given time. By the Rev. JAMES CUMMING, V.P.R.S., F.G.S., Professor of Chemistry, Cambridge.

It has appeared to Professor Cumming that the information conveyed to us by the ordinary instruments for measuring the heating power of the sun's rays is, in one respect, imperfect, in as much as these instruments indicate only the momentary energy of the rays: he was therefore led to devise process which should measure the total result of their action in a given time. The process employed is to expose to the sun a retort with a blackened bulb containing æther, and to note the quantities of this liquid distilled over in different days. In some cases, a second bulb of plain glass has been used to increase the condensing surface, and the apparatus has been otherwise modified. With instruments on this plan Professor Cumming has registered the daily effects of the sun's radiation for more than two years, and he hopes soon to publish his results in a connected form.

On some Electro-magnetic Instruments. By the Rev. JAMES CUMMING, V.P.R.S., Professor of Chemistry, Cambridge.

The instruments exhibited and explained by Professor Cumming consisted of:

1. A galvanometer of four spirals, similar to that described in his translation of Demonferrand, (pl. v. fig. 86,) but formed of flattened copper wire with silk ribbon interposed, each spiral being fixed upon a graduated slide.

2. A Breguet's thermometer, with a conducting wire passed through its axis, for the purpose of measuring either the heat evolved by different galvanic arrangements in passing through a given wire, or that evolved in different wires by the same battery.

On the Thermostat, or Heat-governor, a self-acting physical Apparatus for regulating Temperature. By ANDREW

URE, M.D., F.R.S., &c.

This instrument acts by the unequal expansion of different metals in combination: it admits of many modifications of external form, but, in all, the metallic bars must possess such force of flexure in heating or cooling as to enable their working rods or levers to open or shut valves, stopcocks, and ventilating orifices.

Steel and zinc are the two metals employed: they possess a great difference of expansiveness, nearly as two to five, and are sufficiently cheap to enter into the composition of thermostatic apparatus; but zinc has in reference to the present object one property which should be corrected. After being many times heated and cooled, a rod of that metal remains permanently elongated. This property may, however, be in a great measure destroyed, and considerable rigidity acquired by alloying it with four or five per cent. of copper and one of tin. Such an alloy is hard, close-grained, elastic, and very expansible, and therefore suits pretty well for making the more expansible bar of a thermostat.

Let a bar of zinc or of this alloy be cast, about an inch in breadth, one quarter of an inch thick, and two feet long, and let it be firmly and closely riveted along its face to the face of a similar bar of steel of about one third the thickness. The product of the rigidity and strength of each bar should be nearly the same, so that the texture of each may pretty equally resist the strains of flexure. Having provided a dozen such compound bars, let them be united in pairs by a hinge-joint at each of their ends, having the steel bars inwards. At ordinary temperatures the steel plates of such a pair of compound bars will be parallel and nearly in contact, but when heated they will bend outwards, receding from each other at their middle parts, like two bows tied together at their ends. Supposing this recession to be one inch for 180° Fahrenheit, then six such pairs of bows, connected together in an open frame with rabbeted end plates, and with a guide rod playing through a hole in the centre of each, will produce an effective aggregate motion of six inches, being half an inch for every 15° Fahr., or 84° C. Instead of limiting ourselves to half a dozen such pairs of compound bars, we may readily lodge in a slender iron frame a score or two of them, so as to furnish as great a range of motion as can be desired for most purposes of heat regu