deviation g d L, of the ray La b c refracted by the prism, may be ascertained by mounting the prism with its glasses in front of the object-glass of the telescope of a theodolite, or other angular instrument, as represented in the figures. The prism being so placed as only to cover part of the object-glass, one half is appropriated to the reception of the refracted, and the other to that of the reflected rays; and the telescope being turned from the object, until the images produced by the refracted and reflected rays appear successively in the field, the corresponding deviations are easily ascertained. If the surfaces of the prism are sufficiently flat and well polished to afford a distinct reflected image, a better arrangement for obtaining the angle of the prism, which dispenses with the use of the glass D C, is represented in figs. 4 and 5. The apparatus by which the prism is mounted before the telescope, is represented in fig. 6. R is a collar fitted to the end of the telescope T U, to which it may be fastened by the clamping-screw T. From opposite extremities of the diameter of this collar, two arms N N' project in the direction of the axis of the telescope, which are furnished at their extremities with the screws M M', whose points are presented to each other. By means of the motion which the collar R has on the telescope, and by varying the length of the arm N', which consists of two pieces fastened together with screws S, tapped into one piece, and moving in an elongated aperture in the other, the line joining the points of the screws M M', may be made perpendicular to the limb of the instrument on which the angles are to be measured. The screws M M' carry, between their points, a ring G H, whose plane is therefore perpendicular to the limb of the instrument; and which revolves freely round the line joining the points of the screws. The object of this motion is to adjust the prism to its position of minimum deviation; and for very nice observations, a slow motion may be given to the ring by the wheel and pinion W V, but this may be dispensed with in ordinary cases. Another ring E F, furnished with a piece of perfectly parallel plateglass A B, is fitted into the ring G H, so as to revolve about its centre in the same plane with GH; and to FE are attached the pieces K I, furnished with the screws II, which carry the glass CD. This glass has conical holes drilled in it, for the reception of the points of the screws II; and having a free motion about the line joining their points, it accommodates itself to the surface of any prism that may be placed between it and the glass A B. The prism, having its faces slightly moistened, is thrust, like a wedge, between the glasses until both are found to be in accurate contact with its surfaces. This is ascertained to be the case by the fluid spreading uniformly between the prism and the glasses, and also by the colours produced by the thin plate of intervening air or fluid. When an accurate contact has been obtained, the prism will retain its place firmly, provided too much fluid has not been used; but if any error be apprehended from its shifting, a little soft wax applied at its base, to the glass A B, will prevent it from altering its position. It will be sufficient to explain the use of this apparatus with the theodolite, in order to shew its applicability to any other angular instrument of the same kind; and the manner of using it with the sextant is so perfectly analogous to the method of observation already described, as to render any other description superfluous. In using the theodolite, it is first necessary to make the line of collimation of the telescope parallel to the horizontal limb, which is then adjusted to the plane of the object to be observed, either by raising or lowering the object, or by adjusting the plane of the limb by the parallel plate-screws. The apparatus having been attached to the telescope, the outer ring G H is inclined at a small angle to the line of collimation, and the telescope is turned aside until the image of the object is seen by reflexion from the back of the glass A B. The reflected image is then brought into coincidence with the horizontal wire in the focus of the eye-piece, by turning the collar R upon the telescope, and the collar is made fast by the screw T. The telescope is now pointed 90° from the object, and the ring G H is turned until the reflected image is again visible. If it is now above or below the horizontal wire, it is once more made to coincide with it, by altering the length of the arm N'. The latter adjustment is rendered permanent by tightening the screws S. If the apparatus be in good adjustment, the reflected image of the luminous object should remain on the horizontal wire, at whatever angle the telescope is inclined to the object; and the process of adjustment ought to be repeated until this is the case, It will afterwards be generally sufficient to adjust the instrument by means of the collar R. The apparatus being adjusted, the principal section of the prism, and consequently the plane of reflexion between the glasses, is brought into coincidence with the plane of the divided limb by means of the ring E F. The refracted image will now be upon the horizontal wire of the telescope; and by moving the ring G H, and following the motion of the refracted image, a point will be found where it stops and begins to move in the opposite direction. By means of the tangent screw, the cross-wires are made to intersect the image at this point, at which the deviation is a minimum, and the angle is read off. The telescope is then turned round, until the reflected image intersects the cross-wires, and the angle is read off. The direct bearing of the object being then taken, the differences between it and the other observed angles will give the deviations of the reflected and refracted rays, which must be corrected for the parallax due to the distance of the prism from the centre of the instrument.* From these angles the index of refraction is easily calculated. Let the observed deviation of the refracted rays; =the angle of the prism; and g=the observed deviation of the reflected rays. Then, since the prism is in its position of minimum deviasin(+); which expressed in terms of the observed tion, μ = sin * If the distance of the observed object be great, the correction, in seconds, to be added to an observed angle & will be nearly r sin α where d is the distance of the observed object, and r the distance of the prism from the centre of the theo doli e. It is obvious, that this method of determining refractive powers is particularly well adapted for fluid substances. The fluid may either be contained in a hollow prism made of parallel plate glass, or a few drops may be retained by capillary attraction between the glasses of the instrument; in either case, the angle of the fluid prism will be ascertained in the manner already shown. From a considerable number of observations, I have found that the method which has now been described, affords considerable facilities in determining refractive powers. The construction of the apparatus makes the adjustment of the surfaces of the prism quite independent of each other. For the rings E F, G H (fig. 6), and the face of the prism in contact with the glass A B, remain constantly perpendicular to the divided limb; and the face in contact with the glass CD, is directly adjusted by the motion of the ring EF. The application of glass-plates to the surfaces of the prism also greatly facilitates the process, inasmuch as it renders it unnecessary to have the faces of the prism so highly polished as would otherwise be indispensably requisite. In fact, it is generally sufficient to grind the surfaces smooth with fine emery, when the fluid by which the prism is retained between the glasses will render it perfectly diaphanous. Sir David Brewster, who seems to have been the first to avail himself of this method of making prisms, observes, that, in measuring the refractive and dispersive powers of bodies that are incapable of receiving a good polish, "by cementing upon the two refracting planes pieces of parallel glass with a fluid of nearly the same refractive density, substances like horn, rock salt, and several of the gums, may be rendered perfectly transparent."* In alluding to the circumstance, that the prism is included between two plates of glass, by whose inclination its refracting angle is ascertained, I am led to anticipate some objections to the accuracy of the process that may suggest themselves to the reader. It may be thought, that not only will the refractive power of the glass-plates vitiate the result, but also that a source of *Treatise on New Philosophical Instruments, p. 279. error will arise from the glasses not applying with sufficient accuracy to the faces of prism. No appreciable error need be apprehended from the refractive power of the glasses, provided they be of the kind used for making the mirrors of sextants; and should any notable imperfection exist, it will manifest itself by the separation of the images reflected from the two surfaces of the defective glass. With reference to the other source of error, I may refer to the results of actual observation, to shew within what narrow limits it may be confined; and in doing so, I have to acknowledge the kindness of Mr William Nicol, Mr John Adie, and Mr Alexander Bryson, to whom I am indebted for the use of most of the prisms whose angles have been determined by this method. The angle of a flint-glass prism, belonging to Mr Adie, was taken five times, and the greatest and least result differed by only 50", the object observed being a street lamp nearly 400 feet distant. The angle of the same prism was then determined by a double observation on the turrets of Trinity Chapel, Deanbridge, at a distance of about of a mile, and the result differed from the mean of several accurate experiments made some years ago by Mr Adie, by 30". A plate-glass prism, examined in the same way, gave a result differing from Mr Adie's by 50", and the refractive indices of these prisms differed from Mr Adie's results only in the fourth place of decimals. A similar agreement exists between many other observations I have made on different substances and the results obtained by former observers, so that I feel quite satisfied with the practical efficiency of the process. EDINBURGH, 4 DUKE STREET, 10th June 1843. On Solar Radiation. The experiments mentioned in the text (page 215), referred to a curious inquiry which has occupied my attention for some years, namely, the loss of force which the sun's rays experience in passing through the earth's atmosphere. It might VOL. XXXVI. NO. LXXI.-JAN. 1844 H |