to them others which he deemed necessary for completing the examination of the question.

He contests first the fundamental proposition of the preceding theory, remarking, that if the shadow is the local absence of light, the light may be partial or total, without this necessarily producing any change in the nature of the shadow in question. The interception of the light can only produce shade, but not coloured shade, at least unless the colouring be imparted to it from some other source.

The experiment which we have related, in which black shadows are seen to form in a light decomposed by the prism, or coloured in any other manner, provided the chamber be dark, furnishes another argument against this hypothesis. Mr Trechsel has obtained nearly the same results as M. Zschokke, with reference to this class of shadows. He has made his experiments by passing the light of the sun or of a candle, through disks of coloured glass*, as well in a lightened apartment as in a large drawing camera obscura, in which the object-glass was replaced by differently coloured glasses, in such a manner as to produce the tint desired upon a piece of white paper placed in the bottom. In the camera obscura the shadows are always black, if the light be excluded from all parts; they immediately become coloured when some other light is allowed to penetrate, and the tints which they then assume are always complementary of those of the light transmitted; thus in the red light the shadow is blue or greenish, in the green light it is pale red, &c.

These observations naturally lead us to conclude, that the colouring of shadows does not depend upon the nature of the intercepted light, but rather upon the day-light which mingles with them. This conclusion is enforced by an experiment which does not appear to have been hitherto made, and which is of great weight in the question. If the day light, introduced with proper management, for example, by raising the curtain a little, be made to fall upon the bottom of the camera obscura, when it is coloured green by an object-glass of that colour, the place

The author had for this purpose large squares of coloured glass furnished by the brothers Müller, young artists known by their success in the attempts which they have made, especially of late, to rediscover the art of painting glass, which had been lost for several centuries.

shone upon by the light will assume a pale red colour, without there being any shadow; if the object-glass is red, the light will make the place where it falls appear of a greenish blue.

The following facts come also in support of Mr Trechsel's fundamental opinion.

1st, The natural bluish shadows are more distinctly observed in winter, with an overcast sky and a hazy atmosphere. Now, in these circumstances, the blue light predominates, on account of its greater refrangibility. According to Mr Zschokke's hypothesis, the shadows projected in this blue light would not be blue, but red or orange; it is seen, on the contrary, that the blue tint which they really have, comes from the reflection of the predominant blue light of the day.

2d, If this bluish shadow be illuminated by the yellowish light of a burning candle, it assumes, at the very moment, a yellowish tint.

3d, On the other hand, the black or greyish shadow of the light of a candle assumes a blue colour, whenever some rays of the light of day are made to fall upon it. This is Rumford's experiment.

4th, If there be placed behind the shadow projected by this day light when it is weak, an object painted red, yellow, or any other colour, the shadow immediately assumes a tint similar to that of the object which sent reflected light to it.

5th, The shadows coming from the interception of the light of a candle, are always of a more or less deep black, provided there has been only one candle burning; they appear yellowish when two are lighted, of which the one shines upon the shadow produced by the interception of the light of the other.

"It is easily seen," says Mr Trechsel," that there can be no question here of the inflection of the light in the shadow, either according to the ordinary explanation of this phenomenon, or according to Mr Fresnel's theory of interferences; for, 1st, The coloured shadows are homogeneous, and not composed of alternate bands; 2dly, They are obtained of any breadth that one pleases; 3dly, They preserve, in general, the same colour, although they change their intensity at each variation produced in the distances which separate the plane that receives the light, the opaque body and the source of light."

Mr Trechsel, in consequence, proposes to distinguish two sorts of coloured shadows, one of which may be termed objective, and the other subjective. Among the former would range themselves, 1st, The shadows, whose bluish colouring is owing to the reflection of the day light; 2d, The shadows that are coJoured yellow by the direct light of a candle; 3d, Those which are obtained from the reflection of the light by a neighbouring coloured body. To the subjective shadows would be referred those which are produced in the light coloured either by prismatic decomposition, or by its transmission through coloured glass. In this latter class would also be placed the remarkable phenomenon of the coloration by day-light in the camera obscura, and some other similar phenomena.

The shadows, whose colouring is produced by direct or objective means, do not require further explanation; but the case is different with those whose colouring is only subjective. "With regard to these latter," says Mr Trechsel," Mr Grotthuss's hypothesis appears to me the most probable. It accords not only with ordinary observation, but also with the experiment of the camera obscura which has been described above, and which was not known to Mr Grotthuss. According to this author, when our eye receives the impression of any colour whatever, for example, orange light, transmitted in large quantity, the sensibility of the organ for this light is diminished, and perhaps the sensibility for the complementary blue light increases. If we now make the day light, or any other white light, fall upon a shadow projected in this coloured light, or simply upon a ground tinged with this same light, the orange ray disappears subjectively of the day-light, and we then only perceive the united sensation of the other rays contained in the fasciculus, rays which, by their combination, produce a greenish blue tint, complementary of the orange in the scale of Newton.

No doubt can be entertained of the subjectiveness of the phenomenon of the camera obscura, which I have already several times mentioned, if it be brought to mind that the day-light sometimes produces the red tint in it, sometimes the green, according to the colouring of the ground. Another experiment may be added, which, although not new, is yet not the less striking. Let two candles be placed, so as that two shadows

may be projected from the same opaque body, and the shadow formed in the light of one of the candles be lighted by that of the other. These shadows, as is known, will both be yellowish. Let one of the lights now be coloured red, by making it pass through a plate of glass of that colour; the shadow coming from the interception of the other light, will immediately assume a red tint, (objective colouring); but, at the very instant, the other shadow, which is only shone upon by the pure light of the other candle, will become green, (subjective colouring, produced in the organ of the observer itself, from the defect of the perception of the red ray); and vice versa, if one of the shadows is objectively and directly coloured green, the other will be subjectively coloured red.

Mr Trechsel here remarks, that the phenomenon observed by several members of the Helvetic Society of Natural Science, in a chapel near Soleure, appears to be of an analogous nature to the subjective coloration of shadows, in particular to the pheno menon last cited, and to that of the colouring of the bottom of the camera obscura by the light of the day. It will not be useless to recall here the description of this phenomenon, such as we have already given it in our account of the tenth session of the Helvetic Society *. All the panes of the windows of the chapel, without exception, are of pale yellow glass; the frames of these windows, which are of iron, are perforated here and there with small holes of about a line in diameter; the light which penetrates by these holes, is of the most beautiful blue, even when through them the view is carried upon perfectly white clouds. The same effect is also produced when one of the windows is opened, and the slit thus formed is blue until the opening attains a certain width. We had explained this phenomenon, as probably arising from the psychological effect of contrast. Now this effect, which in general cannot be contested, may be owing to the momentary paralysation in the organ of the faculty of perceiving one of the partial sensations which compose the total impression.

Mr Trechsel concludes his memoir with the following brief review:

* See Bibl. Univers. tom. xxix. p. 326.

"1. Coloured shadows may be distinguished into objective and subjective.

"2. The former owe their colouring to the light which arrives at them either directly or by reflection; they are not therefore total shadows, but are rather, to use the scientific term, penumbra.

"3. The shadows whose colouring is subjective, are the effect of a particular disposition of our organ, which, when it is fatigued by the impression of a single colour, no longer perceives that ray in a fasciculus of white light; so that the complementary ray predominates and communicates its tint to the shadow projected in the primitive light.

"4. So far as we have been able to observe, the eye follows in this process the scale of Newton. If the corresponding colours are not always exactly complementary, it must be attributed to the difficulty of obtaining artificial tints so pure as those of the solar spectrum.

"5. There follows from this, that the colouring of shadows is impossible, if there be no other light than that by the interception of which the shadow is formed. The presence of a light coming from another part, for example, from the sky, or the clouds, is an indispensable condition to the formation of coloured shadows.

"6. Lastly, the shadow is not necessary to make the complementary colours appear. A small quantity of white light, put in prominent contrast with a large mass of coloured light, assumes, in certain circumstances which we cannot well determine, the complementary tint corresponding to the colour of this latter light."

Mr Trechsel's explanation appears to us satisfactory; it introduces, it is true, two causes for a phenomenon which has usually been considered as one; but this is not the first case where a careful analysis has obliged us to admit several agents in an effect single in appearance. Without doubt, in the number of the very varied experiments which may be made on the subject of coloured shadows, there will still present themselves many details which will not be immediately explained; but it is probable that their origin will be found in the peculiar circumstances of