the method which appears to me to fucceed the beft is, to cut a bottle circularly in a fpiral flip of a few lines in breadth. It is very caly to cut a bottle in fuch manner as to form a fingle long flip, and thus unneceffary joinings are avoided.

The whole p is to be plunged into ether, until it is fufficiently foftened, which comes to pafs fooner or later, according to the quality of the vitriolic ether that is employed. Half an hour frequently fuffices; but I have already obferved, that there is a great diverfity in the manner in which different forts of vitriolic ether act, and of which the caufe is not yet, fo far as I know, determined.

The flip being taken out, one of the extremities is to be taken hold of and rolled, firft upon itself at the bottom of the tube, preffing it; then the rolling is to be continued, mounting fpirally along the mould, and taking care to lay over and comprefs with the hand every edge, one against the other, fo that there may not be any vacant fpace, and that all the edges may join exactly. The whole is then to be bound hard with a tape of an inch in width, taking care to turn it the fame way with the flip of elaftic gum. The tape is to be tied over with packthread, fo that by every turn of the packthread joining another, an equal preffure is given to every part: it is then left to dry, and the tube is made.

The bandage is to be taken off with great care, that none of the outward furface, which may have been lodged within the hollows of the tape (of which the caoutchouc takes the exact impreffior) may be pulled away. I advife the application of a tape before the packiliread,

because, especially in the thinner tubes, we fhould run the risk of cutting the caoutchouc, if the pack thread were applied immediately upon it.

It is eafy to take off the tube of elaftic gum which has been formed upon a folid mould of one piece; if the mould be made rather conic, it may be made to flide off by the fmaller end: at the worst, it is eafily accomplished by plunging it into hot water, for it is foftened by the heat, and is diftended; without this precaution it would be fometimes difficult to draw it off when dry, becaufe, having been applied upon the mould whilft it had its vo lume augmented by the interpofition of the ether, the parts of the caoutchouc are drawn nearer cach other by the evaporation of the interpofed body.

The great affinity between these two bodies is feen by the length of time that the odour of the ether remains, notwithstanding the great volatility of the latter, and that the apparent drynefs of the tube seems to fhew that there is none remaining; nevertheless, after a certain time, the odour difappears intirely. One of thofe tubes which are now before you, and which was made with ether, after the method here defcribed, does not retain the leath trace of the folvent. It is needles to say, that it is easy to make tubes as thin, or as thick, as may be judged proper.

Although the procefs that I am now defcribing is but very little expenfive, yet I have tried to employ other folvents in lieu of ether, because it is not to be had in every place, and requires particular care in its prefervation. I have employ ed, with fome fuccefs, the effen

tial oils of lavender, and of turpentine; both of them fpeedily dilate the caoutchouc, and are of no great price. The difagreeable fmell of the oil of turpentine becomes perhaps, in procefs of time, lefs difagreeable than that of lavender. This laft is dearer, but the difference is not fo great as it appears at firft, for we may make fome advantage of the oil of lavender that is employed, by the following operation; upon plunging into alcohol the claftic tube prepared with oil of lavender, the alcohol charges itself with the oil, and forms a very good lavender-water; the fame as would be made by an immediate mixture of oil of lavender with (pirit of wine. Immerfion in this liquor alfo ferves to haften the drying of the caoutchouc inftruments, thus made by means of effential oils. I have made tubes with the oils of turpentine and of lavender; both are much flower in evaporating than ether. The oil of turpentine particularly appeared to me always to leave a kind of stickinefs, and I know not as yet, that we have any means whereby to get speedily rid of its fmell.

Nevertheless there is a folvent which has not that inconvenience; it is cheaper, and may eafily be procured by every one; this folvent is water. I conceive it will appear ftrange to mention water as a folvent of elaftic gum, that liquid having been always fuppofed to have no action upon it. I myself refifted the idea; but, reflecting that ether by being faturated with water is the better enabled to act on caoutchouc, and that this gum when plunged into boiling water becomes more transparent at the edges, I prefumed that this effect was not VOL. XXXVI.

due fimply to the dilatation of its volume by the heat. I thought that, at that temperature, fome action might take place, and that a long continued ebullition might produce more fenfible effects. I was not difappointed in my expectations, and one of thofe tubes was prepared without any other folvent than water and heat. I proceeded in the fame manner as with ether; the elaftic gum dilates but very little in boiling water, it becomes whitish, but recovers its colour again, by drying it in the air and light. It is fufficiently prepared for ufe when it has been a quarter of an hour in boiling water; by this time its edges are fomewhat transparent. It is to be turned fpirally round the mould, in the manner we defcribed before, and replunged frequently into the boiling water, during the time that is employed in forming the tube, to the end that the edges may be difpofed to unite together. When the whole is bound with packthread, it is to be kept fome hours in boiling water, after which it is to be dried, ftill keeping on the binding.

If we wish to be more certain that the connection is perfect, the fpiral may be doubled; but we must always avoid placing the exterior furfaces, of the flips one upon the other, as thofe furfaces are the parts which moft refift the action of folvents. This precaution is lefs neceflary when ether is employed, on account of its great action upon the

is only at a very high temperature that water exercifes any fenfible action upon caoutchouc. I can affirm that at the hundred and twentieth degree of Reaumur's thermometer, [302 of Fahrenheit] this affinity is not fuch as that the water can give a liquid form to caoutchouc; and it does not appear that we have any thing to fear in practice from a combination between these two bodies, which, though it really is a true folution, does not take place in any fenfible degree but at a high temperature. It is therefore, at prefent, eafy to make of caoutchouc whatever inftruments it may be advantageous to have of a flexible, fupple, and elaftic, fubftance, which is impermeable to water at the temperature of our atmosphere, and refifts the action of acids, as well as that of moft other folvents. As to the durability of thefe inftruments, few fubflances promife more than this, because it may be foldered afresh in a damaged part. Any woven fubftance may be covered with it, it is only required that the fubftance fhould be of a nature not to be acted upon, during the preparation, either by ether or by boiling water; for thefe two agents are thole which appear to me to merit the preference. Artifts will frequently find an advantage in employing ether, as it requires lefs time; fo that a perfon may make, in a fingle day, any tube he may have occafion for. The expence of ether is very little, fince it is needful only to difpofe the caoutchouc to adhere; and, being brought into that ftate, the caoutchouc may be kept in a vefiel perfectly well clofed. It would alfo diminish the expence of the ether, if, inftead of washing it with a large

equalities of folar days,. mean interval, or day, has bee found, and divided, by very gener confent, into eighty-fix thoufant four hundred equal parts.

A pendulum, vibrating in m and equal arcs, may be fo adju in its length, as, by its vibrations, t make this divifion of the eart motion into eighty-fix thouland for hundred equal parts, called feconda of mean time.

Such a pendulum then becom itself a measure of determinate length, to which all others may be referred, as to a standard.

But even the pendulum is not without its uncertainties.

Firft, the difficulty of afcertais ing in practice its centre of of lation, as depending on the for of the bob, and its diftance from the point of fufpenfion; and the effect of the weight of the fulpen ing wire, towards difplacing the centre of ofcillation, (that centre being feated within the body of the bob, and therefore inacceffible to measurement), are fources of connderable uncertainty.

Secondly, both theory and exp rience prove that, to preferve in ifochronifm, it must be fhorter 1wards the equator, and longer wards the poles.

Thirdly, to continue fmall d equal vibrations, through a f

ient length of time, and to count efe vibrations, machinery and a ower are neceflary, which may xert a fmall but conftant effort to enew the wafte of motion; and he difficulty is, fo to apply these as Fat they fhall neither retard nor acelerate the vibrations.

In order to avoid the uncertaines which refpect the centre of of illation, I have fubftituted an uniorm cylindrical rod, without a bob, or the pendulum.

Could the diameter of fuch a rod e infinitely fmall, the centre of of illation would be exactly at twohirds of the whole length, meared from the point of fufpenfion. iving it a diameter which fhall nder it fufficiently inflexible, the ntre will be difplaced indeed, it, in a fecond rod, not the 100,000th part of its length, and ot the hundredth part fo much as a fecond pendulum with a bob. is difplacement is fo infinitely nute then, that we may confider e centre of ofcillation, for all actical purposes, as refiding at o-thirds of the length from the int of fufpenfion. The diftance tween these two centres might be fily and accurately afcertained in actice.

But the whole rod is better for a ndard than any portion of it, beufe fenfibly defined at both its tremities.

The uncertainty arifing from the ference of length requifite for the ond pendulum or rod, in differt latitudes, may be avoided by ing on fome one latitude, to which rftandard fhall refer.

To give an uniform impulfe to rod, to renew the wafte of mo1 occafioned by the friction and ftance of the air, I have made

an improvement on the fcapement of the common clock, which effectually prevents any irregularity, that might otherwife be occafioned by the unequal action of the wheels.

To make the experiment, nothing more is neceffary than fuch a clock, and a piece of ftraight iron wire, of about the diameter of a common goofe-quill; and the length that will be found to vibrate feconds, will be about fifty-eight English inches, from the point of fufpenfion to the end of the rod. The most certain way of proceeding in the experiment is, to have the rod at firft too long, and to continue cutting fmall portions off the lower end, till the clock is brought to time.

The advantage that the above has over all other pendulums, for afcertaining an univerfal ftandard of meafure, are,

Firft, that it can be regulated with more certainty.

Secondly, that it is one-third longer, and therefore can be divided with more certainty.

And, thirdly, its fimplicity, which puts it in the power of any perfon to make the experiment with the greateft accuracy, even without the ufe of figures; the whole (after the clock is regulated) being afcertained by one fingle measurement of a ftraight line.