be sensibly achromatic throughout.* This, with a source of light of small magnitude, where the boundary of the illuminated sea area will be sharply defined, and where the effects of chromatic dispersion are likely to be otherwise quite appreciable, is a valuable property.

Indeed, it may be worth inquiry whether lenticular apparatus for the electric light might not advantageously be rendered, at least approximately, achromatic; or failing this, whether achromatic esoptric zones should not be alone employed to the exclusion of lenticular agents.

Panoptic Apparatus with Acryptic Prisms.

60. The panoptic arrangement, which permits the middle back rays, or rays whose deviations approach to 180°, to be transmitted through interspaces left between the front prisms, may also be employed for the transmission of the marginal back rays, that is, rays with deviations approaching to, but not greatly exceeding 90°.

In fig. 16 is represented an instrument in which the middle back rays are first reflected through the flame, and then pass through the front prisms and lenticular apparatus as in Mr Stevenson's holophotes, while the marginal back rays are reflected through interspaces left between the outer front zones by acryptic back zones, whose sections are of the forms described in article 40 or article 42.

An arrangement is thus obtained in which only one fourth of the whole light suffers successive transmission through two optical agents, namely, the back mirror and front prisms or lenticular apparatus, while three-fourths of the light is sent directly to the eye by single optical agents.

61. I need scarcely add, that were it wished, the whole of the back light, by help of the prisms described in this paper, could be transmitted on the panoptic principle

* In some cases it might be desirable to make the back prisms next the axis diesoptric, and those more remote triesoptric. We should thus render achromatic those prisms in which the effects of chromatic aberration would otherwise be most hurtful, and, at the same time, avoid employing three reflections where they might not be much needed.

through a sufficient number of interspaces left between the front prisms.

Fig. 16.

Fixed Light Apparatus with Acryptic Prisms.

62. If the sections of the prisms in figure 16 revolve through 180°, about a vertical axis passing through the luminous origin, they will obviously generate a series of horizontal zones capable of condensing all the light of a lamp, so as to illuminate one-half of the horizon. The arrangement figured, having been originally designed, as already explained, for another purpose, the uppermost and lowermost prism profiles must be omitted as redundant, when it represents the section of a fixed light apparatus.

540

Some New Arrangements for Lighthouse Illumination, including those of the Tay Leading Lights. By THOMAS STEVENSON, F.R.S.E., M.I.C.E.* (With Three Plates.)

Since the adoption of the holophotal system of illuminating lighthouses, several modifications and improvements have occurred to me in designing apparatus suited for particular localities. Some of these designs have been carried into execution, while others have not yet been adopted. I have thought it right to describe the more important of these designs, as other engineers may find occasion to employ them.

Improved Form of Holophote and of Azimuthal Condensing Apparatus

In a treatise on the Design and Construction of Harbours, published in 1864, I described a form of holophote and fixed light apparatus, which possessed considerable advantages over those formerly described. If a spherical mirror, subtending even an angle of only 100°, be used for returning the back rays through the flame, it is obvious that there must necessarily be a considerable loss of light from two causes. First, because two agents are employed; and, secondly, because the light which passes upwards from the flame near the vertical axis, being returned in the contrary direction, must impinge upon the metallic burner and be lost. In order to obviate entirely these disadvantages, the following arrangements were proposed:

In the case of the holophote, as represented, figs. 1 and 2, the front rays are acted on by a half holophote, a, b, C, subtending 180°, by which single agent the whole of the front hemisphere of rays is at once parallelised. The spherical zone of say 60°, reckoning backwards from the vertical axis, is next parallelised by the single agency of a paraboloidal

*Communicated 6th December 1867, and read before the Society 27th January 1868. Awarded, jointly with Professor Swan's communication ou New Forms of Lighthouse Apparatus, the Keith Prize, value Forty Sovereigns.

mirror, h, e, f, i, or by paraboloidal strips. The balance of the back light, amounting to 60°, is returned by means of the dioptric spherical mirror e, f, so as to pass through the flame and be ultimately parallelised by the half holophote which is placed in front of it.

By means of this arrangement, it will be seen that only 60° of the whole sphere pass through more than one agent, leaving the remaining 300° to be parallelised by only one.

In a similar manner, the rays in a fixed light of 180° in azimuth were acted on. In figs. 3 and 4 the front half of the light is at once parallelised in the vertical plane by a half of Fresnel's beehive apparatus, a, b, c, while 26° of the back rays

were parallelised in the vertical plane by paraboloidal strips, g, h, i, j, which form a series of truncated paraboloidal domes, and the balance, or 128°, was returned through the flame by a sector of the dioptric spherical mirror e, f.

I also stated in my book on Harbours, that reflecting prisms of glass might be employed for the back rays in place

of metallic paraboloids; and in the Minutes of the Institution of Civil Engineers for 1867, I stated that those prisms could not be formed of "glass of the ordinary refractive index." But though prisms of Fresnel's profile could not be used, it was found on trial that prisms of ordinary glass of the form represented in fig. 5, could be employed as far back as 180°. And here I would remark that my assistant, Mr Brebner, C.E., whose optical protractor was used, is, I consider, entitled to equal credit with myself in the matter. Fig. 5 represents in vertical section a holophote, a, b, c, with the new form of back reflecting prisms, a, g, and c, h. The dioptric mirror is shown by letters i, j, k. It is further satisfactory to know, that Mr Chance, to whom I communicated a drawing of them on 27th September 1867, wrote to me, "I cannot see any real difficulty in the execution. of your ingenious form of prism for deflecting the back light."*

It is further worthy of remark, that the designs which have just been described may be found useful for azimuthal condensing lights, whether catoptric or dioptric, especially when made of a moderate size. For example, we could dispense with the upper double agents, which were used in the Buddonness apparatus, about to be described, by adopting strips of metallic, truncated, paraboloidal domes, or by domes of the dioptric prisms just described, subtending, in plano, the required angle.

New Condensing Lights at Buddonness for 45° in Azimuth. The Trinity-House of Dundee having employed Messrs Stevenson to design new leading lights for Buddonness, at

* I very willingly add, that my friend Professor Swan of St Andrews has also independently devised a similar form of prisms. A drawing of my design, as stated above, was sent to Mr Chance on the 27th September, and my paper was sent to the Secretary on 6th December; and Professor Swan, unaware of what I had done, communicated his on the 9th December. I must, however, refer my readers to Professor Swan's own paper for many other ingenious arrangements, and, in particular, for his curious method of dispensing with the spherical mirror.