pher having descended in a diving bell to a con siderable depth in the sea, while it was illuminated by the sun, was much surprised to see the back of his hand, which received the direct rays, of a beautiful rose colour, while the lower part, which received the reflected rays, was blue. This indeed is what ought to take place, if we suppose that the rays reflected by the surface of the sea, as well as by the minute parts of the middle of it, are blue rays. In proportion as the light penetrates to a greater depth, it must be more and more deprived of the blue rays, and consequently the remainder must incline to red.

PROBLEM LVII.

Why the Shadows of bodies are sometimes Blue, or Azure coloured, instead of being Black.

It is often observed at sun-rise, during very serene mornings, that the shadows of bodies projected on a white ground, at a small distance, are blue or azure coloured. This phenomenon appears to us to be sufficiently curious to deserve here a place, as well as an explanation,

If the shadow of a body exposed to the sun were absolute, it would be perfectly black, since it would be a complete privation of light; but this does not really take place: for to be so, the field of the heavens ought to be absolutely black; whereas it is blue, or azure coloured, and it is so only because it sends back to us chiefly blue rays, as already observed,

The shadow therefore projected by bodies exposed to the sun, is not a pure shadow, but is itself

illuminated by that whole part of the sky not occu pied by the luminous body. This part of the hea, ven being blue, the shadow is softened by the blue or azure coloured rays, and consequently must appear of that colour. It is exactly in the same manner that in painting, reflections are tinted with the colour of the surrounding bodies. The shadow which we here examine, is nothing else than a shadow mixed with the reflection of a blue body, and therefore it must participate in that colour.

It is well known that this phenomenon was first observed and explained by Buffon.

But it may here be asked, why are not all shadows blue? In reply to this question, we shall observe, that to produce this effect, the concur rence of several circumstances are necessary: Ist, a very pure sky, and of a very dark blue colour; for if the heavens be interspersed with light clouds, the rays reflected from them, falling on the bluish shadow, will destroy its effect; if the blue be weak, as is often the case, the quantity of the blue rays will not be sufficient to enlighten the shadow. 2d, The light of the sun must be livelier than it usually is when that luminary is near the horizon, in order that the shadows may be full and strong. But these circumstances are rarely united. Besides, the sun must be only at a small elevation above the horizon; for even when at a moderate altitude there is too much splendour in the atmosphere, to allow the blue rays to be sensible. This light renders the shadow less strong, but does not tinge it blue.

PROBLEM LIX.

Experiment on Colours.

HOLD before your eyes two glasses of different colours, the one blue suppose, and the other red; and having placed yourself at a proper distance from a candle, if you shut one of your eyes, and look at the light with the other, that for example before which the blue glass is held, the light will appear blue. If you next shut this eye and open the other, the flame will appear red; and if you then open them both, you will see it of a bright violet colour.

Every person almost, in our opinion, must have foreseen the success of this experiment; which we have mentioned, merely because an oculist of Lyons, M. Janin, thought he could deduce from it a particular consequence; which is, that the retina may perform the part of a concave mirror, and reflect the rays of light, so that each eye forms at a certain distance an aerial image of the object. Both eyes forming each an image afterwards in the same place, the result is a double image, one blue and the other red, which by their union produce a violet image, in the same manner as when red and blue rays are mixed together. But this explanation will certainly not bear to be examined according to the true principles of optics. How is it possible to conceive that such an image can be formed by the retina? Is it not more probable, and more agreeable to the well known phenomena of vision, that from the two impressions received by he two eyes, there is produced in the common sen

sorium, or in the place where the optic nerves are united in the brain, one compound impression? In this experiment therefore the same thing must take place, as when a person looks at a candle with one eye, through two glasses, the one red and the other blue. In this case the flame will be seen of a violet colour, and consequently it must have the same appearance in the former.

PROBLEM LX,

Method of constructing a Photophorus, very convenienț to illuminate a table where a person is reading or writing.

CONSTRUCT a cone of tin-plate, 4 inches in diameter at the base, and 7 inches in height, measured on the slant side; which may be easily done by cutting from a circle, of 7 inches radius, a sector of 109 degrees, and bending it into the form of a cone. Then through a point in the axis, 2 inches distant from the summit, cut off the upper part of the cone by a plane inclined to one of its sides at an angle of 45°. The result will be an elongated elliptical section, which must be placed before a candle or other light, as near to it as possible, the plane of the section being vertical, and the greatest diameter in a perpendicular direction. When disposed in this manner, if the flame of the candle or lamp be raised 12 or 13 inches above the plane of the table, you will be astonished to see the vivacity and uniformity of the light which it will project over an extent of 4 or 5 feet in length.

M. Lambert, the inventor of this apparatus, ob. serves that it may be used with great advantage to

give light to those who read in bed; for by placing a lamp or taper furnished with this photophorus upon a pretty high stand, at the distance of 5, 6, or 8 feet, from the bed, it will afford a sufficiency of light without any danger. He says he tried this apparatus also in the street, by placing a lamp furnished with it in a window raised 15 feet above the pavement, and that its effect was so great, that at the distance of 60 feet, a bit of straw could be distinguished much better than by moon light, and that writing could be read at the distance of 35 or 40 feet. A few of these machines, placed on each side of a street, and arranged in a diagonal form, would consequently light it much better than any of the means hitherto employed. See Mémoires de l'Academie de Berlin, ann. 1770.

PROBLEM LXI.

The place of an object, such for example as of a piece of paper on a table, being given; and that of a candle destined to throw light upon it; to determine the height at which the candle must be placed, in order that the object may be illuminated the most possible.

THAT We may exclude from this problem several considerations, which would render the solution of it very difficult, we shall suppose the object destined to be illuminated to be very small, or that it is only required that the middle of it shall be illuminated as much as possible. We shall suppose also that the light is entirely concentrated into one point, where the splendor of all its different parts is united.

But, it is well known that the light diffused by