quisite for explaining the elements of natural philosophy. His views in the latter were sound, often original, and always explained with great clearness and simplicity. The mathematical and experimental parts were so happily combined, that his lectures communicated not only an excellent view of the principles of the science, but much practical knowledge concerning the means by which those principles are embodied in matter, and made palpable to sense.

Mr. Robison, who now succeeded to this chair, had also talents and acquirements of a very high order. The scenes of active life in which he had been early engaged, and in which he had seen the great operations of the nautical and the military art, had been followed or accompanied with much study, so that a thorough knowledge of the principles, as well as the practice, of those arts had been acquired. His knowledge of the mathematics was accurate and extensive, and included, what was at that time rare in this country, a considerable familiarity with the discoveries and inventions of the foreign mathematicians.

In the general outline of his course he did not, however, deviate materially from that which had been sketched by his predecessors, except, I think, in one point of arrangement, by which he passed from dynamics immediately to physical astronomy. The sciences of mechanics, hydrodynamics, astronomy, and optics, together with electricity and magnetism, were the subjects which his lectures embraced. These were given with great fluency and precision of language, and with the introduction of a good deal of mathematical demonstration. His manner was grave and dignified; his views, always ingenious and comprehensive, were full of information, and never more interesting and instructive than when they touched on the history of science. His lectures, however, were often complained of, as difficult and hard to be followed; and this did not, in my opinion, arise from the depth of the mathematical demonstrations, as was sometimes said, but rather from the rapidity of his discourse, which was in general beyond the rate at which accu ate reasoning can be easily followed. The singular facility of his own apprehension made him judge too favourably of the same power in others. To understand his lectures completely was, on account of the rapidity and the uniform flow of his discourse, not a very easy task, even for men tolerably familiar with the subject. On this account his lectures were less popular than might have been expected from such a combination of rare talents as the author of them possessed. This was assisted by the small number of experiments he introduced, and à view that he took of natural philosophy which left but a very subordinate place for them to occupy. An experiment, he would very truly observe, does not establish a general proposition, and never can do more than prove a particular fact. Hence he inferred, or seemed to infer, that they are of no great use in establishing the principles of science. This seems an erroneous view. An experiment does but prove a particular fact; but by doing so in a great number of cases it affords the means of

discovering the general principle which is common to all these facts. Even a single experiment may be sufficient to prove a very general fact. When a guinea and a feather, let fall from the top of an exhausted receiver, descend to the bottom of it in the same time, it is very true that this only proves the fact of the equal acceleration of falling bodies in the case of the two substances just named; but who doubts that the conclusion extends to all different degrees of weight, and that the uniform acceleration of falling bodies of every kind may safely be inferred.

A society for the cultivation of literature and science had existed in Edinburgh ever since the year 1739, when, by the advice, and under the direction of Mr. Maclaurin, an association, formed some years before for the improvement of medicine and surgery, enlarged its plan, and assumed the name of the Philosophical Society. This Society, which had at different times reckoned among its members some of the first men of whom this country can boast, had published three volumes of memoirs, under the title of Physical and Literary Essays; the last in 1756, from which time the Society had languished, and its meetings had become less frequent. At the time I am now speaking of it was beginning to revive, and its tendency to do so was not diminished by the acquisition of Mr. Robison, who became a member of it soon after his arrival. It had often occurred that a more regular form, and an incorporation by royal charter, might give more steadiness and vigour to the exertions of this learned body. In 1783, accordingly, under the auspices of the late excellent Principal of this University, a royal charter was obtained, appointing certain persons named in it as a new society, which, as its first act, united to itself the whole of the Philosophical.

Professor Robison, one of those named in the original charter, was immediately appointed Secretary, and continued to discharge the duties of that office till prevented by the state of his health several years after.

The first volume of the Transactions of this Society contains the first paper which Professor Robison submitted to the public, a Determination of the Orbit and the Motion of the Georgium Sidus directly from Observations, read in March, 1786. This planet had been observed by Dr. Herschell on March 13, 1781, and was the first in the long list of discoveries by which that excellent observer has for so many years continued to enrich the science of astronomy. Its great distance from the sun, and the slowness of its angular motion, which last amounts to little more than four degrees from one opposition to the next, made it difficult to determine its orbit with tolerable accuracy, from an arch which did not yet exceed an eighteenth part of the whole orbit. This was an inconvenience which time would remedy; but impatience to arrive even at such an approximation as the facts known will afford is natural in such cases, and Professor Robison, as well as several other mathematicians, were not afraid to attempt the problem, even

in this imperfect state of the data. It is well known that the observations which best serve the purpose of determining the orbit of a planet are those made at its oppositions to the sun, when an observer in the earth and in the sun would refer the planet to the same point in the starry heavens, or when, in the language of astronomers, its heliocentric and geocentric places coincide. Of these oppositions in the case of this planet there were yet only four which had been actually observed. Dr. Herschell had, however, discovered the planet soon after the opposition of 1781 was passed, and though of course that opposition was not seen, yet from the observations that were made so soon after, Professor Robison thought he could deduce the time with sufficient accuracy. The opposition of the winter 1786 he observed himself; for though there was unfortunately no observatory at Edinburgh, he endeavoured to supply that defect on the present occasion by a very simple apparatus, viz. a telescope on an equatorial stand, which served to compare the right ascension and declination of the planet with those of some known stars which it happened to be near. His general solution of the problem is very deserving of praise; and though the method pursued is in its principle the same with all those which ever since the time of Kepler have been employed for finding the elements of a planetary orbit, it appears here in a very simple form, the construction being wholly geometrical, and easily understood. The elements, as he found them, are not very different from those that have since been determined from more numerous and more accurate observations.

When Dr. Herschell first made known this most distant of the planets, many astronomers believed that they had discovered the source of those disturbances in our system which had not yet been explained. Professor Robison was of this number; for he tells us in the beginning of his paper that he had long thought that the irregularities in the motion of Jupiter and Saturn, which had not been explained by the mutual gravitation of the known planets, were to be accounted for by the action of planets of considerable magnitude, beyond the orbit of Saturn. Subsequent inquiry, however, has not verified this conjecture; the irregularities of Jupiter and Saturn have since been fully explained, and are known to arise chiefly from their action on one another, a very small part only being owing to that of the Georgium Sidus, or of any of the other planets.

The next publication of Professor Robison was a paper in the second volume of the same Transactions, On the Motion of Light, as affected by Refracting and Reflecting Substances, which are themselves in Motion.*

The phenomena of the aberration of the fixed stars are well known to depend on the velocity of the earth's motion combined with the velocity of light; the quantity of the aberration, when all

* Edinburgh Transactions, vol. ii. p. 83.

other things are given, being directly as the first, and inversely as the second. It is not, however, the general or the medium velocity with which light traverses space, but it is the particular velocity with which it traverses the tube of the telescope, that determines the quantity of this aberration. Were it possible, therefore, to increase or diminish that velocity, the aberration would be diminished in the first case, and increased in the second. But according to the principles now generally received in optics, the velocity of light is increased when it traverses a denser medium, or one in which the refraction is greater; and therefore were the tube of a telescope to be filled with water instead of air, the aberration would be diminished. Professor Robison, and his friend Mr. Wilson, Professor of Astronomy at Glasgow, had speculated much on this subject, and made many attempts to obtain a water telescope, but hitherto without effect. A paper of Boscovich on the same subject seemed to suggest some new views, that might render the experiment more easy to be made. That philosopher maintained that in ascertaining the effect of a water telescope on the motion of light, the observation of celestial objects might be dispensed with, and that of terrestrial substituted in its place. He argued that while light moves with an uniform velocity, the telescope must be directed, not to the point of space which the object occupied when the particle was sent off which is entering the telescope, but to a point advanced before it by a space just equal to that which both the object and the observer have passed over in the time in which the particle has passed from the object to the eye. It is therefore directed exactly to the place which the object is in when the light from it enters the eye. If, therefore, the ray, on entering the telescope, is made to move faster than it did before, the telescope must not be inclined so much, and the apparent place of the object will fall behind its true place. If the ray is retarded on entering the water, the contrary must happen. Hence a number of very unexpected phenomena would result, affording, without having recourse to the heavenly bodies, a direct proof of the motion of the earth in its orbit, as well as a resolution of the question whether light is accelerated or retarded on passing from a rarer to a denser medium.*

On this reasoning Professor Robison has very well remarked that it would be just if the light, on entering the water telescope, had only its velocity changed, and not its direction. But this is not the case; for the ray that is to go down the axis of the telescope is not perpendicular to the surface of the fluid; it makes an angle with it, depending on the aberration, and therefore in some cases less by 20" than a right angle. On this account the effect is not produced which Boscovich's reasonings lead us to expect.

The sequel of the paper is also full of ingenious remarks.

* Boscovich, Opera Math. tom. ii. opusc. 3.

(To be continued.)

ARTICLE II.

On the Stability of Vessels. By Col. Beaufoy, F.R.S. With Two

MY DEAR SIR,

Plates.

(To Dr. Thomson.)

Bushey Heath, Dec. 22, 1815. THAT part of hydrostatics, which treats on the stability of floating bodies, naturally interests the curiosity of most persons in a maritime country like Great Britain, and excites the desire of many to become acquainted with the law which regulates their equilibrium. Being one of those who are attached to this interesting subject, I take the liberty of laying before you a series of experiments, which, should they prove instrumental in throwing new light on naval architecture, or in improving the construction of vessels, will amply recompense the trouble they cost, in the hope that the time and expense bestowed on them have not been uselessly employed. I remain, my dear Sir, yours very sincerely,

MARK BEAUFOY.

Experiments to verify the Theorems on Stability, particularly M. Bouguer's, with a Description and Drawing of the Apparatus with which they were made, and some Remarks on the Formation of Vessels.

THE principal object in making these experiments was to bring to the test of experiment the different theorems of various writers on naval architecture, particularly those of M. Bouguer for calculating the stability of variously shaped floating bodies. This Gentleman founds his theory on the supposition that the angles of inclination assumed by floating bodies are evanescent, which in a practical sense may be regarded as angles which are very small. But as vessels at sea frequently, by the pressure of their sails, as well as by the action of the waves, make very large angles with the horizon or surface of the water, before implicit confidence, be placed in any theory, it is but prudent to submit it to the test of experiment. The first theorem to be examined is,

That the stability is in proportion to the squares of the areas of the horizontal surfaces or sections.

2. That the height of the metacentre of the parallelopipedon above the centre of gravity of the displaced water is found by dividing the square of half the breadth of the parallelopipedon by three times the draft of water, that is, the depth to which the parallelopipedon is immersed in the fluid.

3. That the metacentre of a right angle triangle is elevated as much above the line of floatation as the centre of gravity of the displaced water is depressed below the surface.

4. That the metacentre of a semi-ellipsis above its line of floata