Our countryman Ferguson, and his cotemporary ge nius, Ben. Martin, ought not to be passed over in silence, as they both contributed largely to explain the construc tion of the different orreries and planetaria, which they themselves improved from time to time, but which re main as specimens only of their ingenuity, since the mo dern discoveries of five additional primary planets has rendered their best labours defective. The improvement of instruments is a natural consequence of discoveries in any art or science ; and it is the fate of those that have long been in use, and much esteemed, to sink in estima tion in proportion as more appropriate substitutes are brought forward to supplant them. While a celebrated mechanist in Paris, Antide Janvier, was employed, in Bonaparte's reign, to construct a planetary machine for the instruction of his son, an instructor of British youth directed his attention to the construction of more simple, but equally accurate machines for explaining the pheno mena of the solar system, as it includes all the planets hi therto discovered. The instructor we allude to is the Rev. W. Pearson, LL.D. and F.R.S. at present the trea surer of the Astronomical Society of London, which is greatly indebted to his co-operation for its existence as an useful society. As we have been favoured by this gentle man's permission to describe some of the planetary ma chines constructed under his superintendence, by the late artist Fidler, we presume we shall render a more accepta ble service to the public, than if we attached greater im portance to more expensive, but less instructive pieces of mechanism, that have survived the estimation in which they were originally held. Of course we do not wish to bring into disrepute those scenic representations of the heavenly bodies, which are produced by moving transpa rencies, of any description, for the amusement rather than the instruction of a wondering audience ; but our aim is to present to our readers, whom we must consider as composing the scientific class of British inhabitants, an account of machinery equally calculated to amuse the learned, and to instruct the learner.
Examination of Planetary Wheel-work-.
Every person who purchases a planetary machine, of any description, ought to make himself acquainted with the powers and properties of that machine; but few per sons would be able to do this without some previous gene ral instruction how to proceed. When the mechanism consists of as many single pairs of wheels as there are heavenly bodies to be actuated, which is the case with Roemer's satellite instrument and planetarium, the first thing to be done is to ascertain in what assumed period the driving, or first moving wheel, is intended by the maker to revolve; which generally may be done by noticing the index and dial, if any, that are connected with the arbor that carries the said first wheel ; but if there he no such index, then the handle itself may be considered as the in dex, and its connexion with the first wheel will show, after a few turns, whether an increase or decrease of velocity is produced in the arm or lever that carries the body, which is usually the last mover. In this way, it will be
readily discovered whether the handle turns in a day, a week, or a year. Let us take Roemer's satellite instru ment for examination, as one of the simple cases: here it is immediately perceived that the first and second satel lites move with a velocity exceeding that of the handle; that the third satellite has apparently the same velocity, or nearly so ; and that the velocity of the fourth is slower than the motion of the handle : hence it may be conclud ed, that, as the third satellite revolves in nearly a week, seven days may be taken as the period of the handle's re volution, from which the periods of the satellites are to be derived by the wheelwork. Indeed this would appear at first sight, if a dial, indicating the seven days of the week, were observed in connexion with the arbor of the handle. In the next place, the teeth of each wheel must be counted carefully, and the wheels must be put down in pairs, as they are observed to act together, in the form of as many fractions, with the driven wheels as the numerators, and the driving wheels as the denominators : then in each pair the numerator will be in the same proportion to the denominator, as seven days (the pe riod assumed for the common arbor of all the first mov ing wheels,) are to the respective periods of the re volving planetary arms. Hence if the assumed period be multiplied by the numerators of each fraction, and divid ed by the denominators respectively, the resulting num bers carried into days, hours, minutes, and seconds, in the usual way, will be the times of the periodic revolutions produced by the wheels in question ; and the times de duced may then be compared with the true periods, as in the subjoined small table.
In this examination it might appear, at first sight, that the error in the synodic period of the fourth satellite is the greatest ; but when it is considered, that the motion of that body is much slower than that of any of the rest, it will be found that +2' 6".25, in the period of the second satellite, is a greater error than —5' 7".09 in that of the fourth ; and accordingly we find among Ferguson's im ,33 32 provements* the substitution ot— 65 for thereby produc ing a period in 13h. 17' 32".3, in which the error was reduced to —0' 21".45. In this machine the wheelwork was contained in a long box, carrying a dial on the cover divided into seven days of the week, to which a hand fix ed on the common arbor of the four driving wheels, con stantly pointed, to indicate the day as the handle revolved; but we do not learn that there was any hour circle, to show the position of the satellites at any given hour of the day.
About the termination of the last century, the Rev. Dr. Pearson contrived a simple machine to answer the same purpose as Roemer's, but without knowing that Roemer had ever made one; and as the assumed period was a so lar day, it will afford a second example for our method of determining the periods produced by the wheclwork.