specularly by the glass and by the surface of the liquid is intercepted, and scarcely any light passes but that diffused by the globules of the blood; with the spectroscope the two black bands characteristic of the spectrum of oxygenated blood are seen most distinctly. The experiment can also be made with filter-paper wetted with blood-but is then less conclusive, because the white fibres of the paper contribute largely to the diffusion.

In quitting this subject, let us observe that, to the explanation of the colour of bodies by fluorescence, objections can be made which it seems difficult to answer. A plate of obsidian glass roughed gives by diffusion a residue of non-polarized light; if it is due to fluorescence, why is it not seen when the obsidian is polished, in the same way as with a plate of uranium glass polished the light emitted by fluorescence is distinguished perfectly? and how is it to be explained that a transparent body (such as quartz) which manifests no fluorescence in its interior, or at its polished surface, can take isochromatic fluorescence when reduced to powder, or simply when its surface is roughed ?

A Study of some Crystallized Substances.

I return now to the question whether a medium perfectly homogeneous and not fluorescent can be illuminated by the passage of a pencil of rays. This point appears to me important, because lateral propagation of light in this case is contrary to what has been generally admitted in the theory of undulations.

Crystallization is a very effective means of purifying substances; and good crystals present a high degree of homogeneity. Still this is by no means absolute; and it is well known that, except in very rare cases, the passage of a pencil of sunlight reveals defects (flaws or the presence of foreign corpuscles) in the interior of crystals that appear perfectly limpid in diffused light. The following are the results I have obtained in studying illumination in some crystallized substances:

Quartz. Of this I studied a great number of specimens *. They exhibit the most diverse properties with respect to illumination; I will mention a few examples of this.

1. Most frequently fine specimens of hyaline quartz, sufficiently pure to be employed in the construction of prisms or other optical apparatus, are not absolutely homogeneous; and when traversed by a pencil of sunlight, phenomena of illumina

* A portion of these belong to the Geneva Museum; but the most remarkable came from a collection of quartz objects acquired at the Vienna Exhibition by M. Rymtowt-Prince, who kindly permitted me to study them.

tion are produced. The trace of the pencil is visible; only it is in general unequal, and more marked in certain places than in others. On examining this trace with a lens or with a microscope, three cases can be distinguished :

(a) Very often the trace is due to little defects of crystallization, to little flaws, which manifest themselves under the appearance of luminous strokes following one or two determinate directions*. Under a vision-angle of 90° the polarization of the light thus diffused is complete, provided the defects of crystallization are very minute; when the flaws are large enough to be readily visible to the naked eye, the polarization is only partial. If the employment of the lens and of the analyzer be combined, these little defects of homogeneity, very apparent when the analyzer is in the position in which it lets pass the light polarized in the plane of vision, totally cease to be visible when the analyzer is rotated into the extinction-position.

(b) Sometimes the trace is due to minute bubbles or cavities which are discovered by means of a lens or a microscope. The light is here again completely polarized; the bubbles become invisible if viewed at a right angle to the pencil, with the analyzer suitably oriented.

(c) In some specimens, or in some parts of a specimen, I have not succeeded in resolving the trace with a low power of the microscope. But I have no doubt that, in this case again, it owes its origin to imperceptible defects or to foreign particles: indeed no limit can be assigned to their minuteness.

2. Some specimens of hyaline quartz, very pure and much rarer, are, so to say, destitute of the power of illumination.

(a) I possess a seal of Siberian quartz which is cut in such a way as very well to permit the observations: its general form is that of a six-sided prism with bases perpendicular to the edges. When placed in the dark room and traversed by a pencil of direct sunlight, no trace is to be seen. On concentrating the light by the process described above for the study of the reflecting-power of flames, taking at the same time the greatest precautions to avoid all false light, a slight trace can with difficulty be distinguished.

(b) Another seal, of Alpine quartz, the shape of which is rather less suitable for the observations, is incontestably illuminated in certain parts; but in others it has been impossible for me to see the trace.

3. A very beautiful seal of yellow Siberian quartz (false topaz), appearing quite limpid in diffused light, is brightly illuminated by a pencil of sunlight; the trace is bluish, and at first sight might be thought due to fluorescence. But on observing it with some care, the light emitted is recognized to be com* I cited a case of this kind in the Archives, 1870, tome xxxvii. p. 155.

pletely polarized, and moreover not equal in every part of the crystal. Examination with the microscope reveals many little flaws in certain parts of the trace; but I did not succeed in resolving the whole of it *.

I found in another seal of false topaz, from Brazil, the same properties, which are probably characteristic of this variety of quartz.

4. In smoky quartz, when it is not too dark, the trace is in general observed, but in very various degrees. When it is feeble, polarization is complete.

In a very fine brown specimen from Siberia, the defects of homogeneity manifest themselves, through transparence to diffused light, under the form of cloudy strata, indicating unequal distribution of the colouring-matter. With sunlight the trace is very marked, very incompletely polarized, and with a lens a multitude of defects and particles are distinguished. This incomplete polarization is a character always recognized when we operate on a liquid holding in suspension particles too numerous or of too great a volume.

5. A small specimen of amethyst cut in the shape of a prism with three sides, and with faces perpendicular to the edges, gave no visible trace in the greater part of the crystal; but in a few points small flaws are observed.

The study of these phenomena in quartz is often complicated by its property of chromatic polarization. Thus, oy making the pencil of light pass, at a little distance from a face, parallel to the axis, and observing the trace with an analyzer, phenomena of coloration or depolarization can be seen, according to the orientation.

Rotatory polarization plays also an important part when the pencil is directed along the axis of the crystal. Here is a remarkable instance :-The second seal of false topaz above mentioned is cut so that its axis coincides with that of the crystal; it presents at its two extremities faces perpendicular to this axis. When a pencil of polarized light enters by one of these two faces to go out at the other, as the substance has a great power of illumination there is reproduced on a small scale the beautiful experiment realized by M. Lallemand when he employed liquids possessing rotatory power. The phenomenon shows itself in coloured fringes which are displaced in the interior of the crystal when the Nicol is rotated that polarizes the incident light. The

* In truth, observation with the microscope is somewhat difficult: the luminous pencil, and consequently the trace, has always too large a section for the whole thickness to be brought to the point; the diffused light with which the entire field of the microscope is washed is opposed to the visibility of very small particles or defects.

first fringes, on the side where the pencil enters, are very bright and distinct; they grow confused and are effaced towards the other extremity.

Calcareous Spar.—I have found, like M. Lalleinand, that fine specimens of Iceland spar present no power of illumination properly so called. The trace of a pencil of sunlight manifests itself solely by the slight red fluorescence of this substance; and no polarization is observed. But if we operate on certain crystals which exhibit minute defects, we can also obtain a white The double refraction of the spar complicates the observation of the polarization-phenomena; yet, with some attention, it has appeared to me that the facts are in this case conformable to what could have been foreseen.

trace.

Rock-salt.-The only good specimen of rock-salt I have had at my disposal is a very fine prism belonging to the Physical Cabinet of the University of Geneva. In it a pencil of light produces a very slight trace, resolvable with the lens into small bubbles or cavities. The polarization is complete; and the defects are visible or disappear according to the orientation of the analyzer. This is the same as I have already stated for quartz.

In the other, less pure specimens which I have examined, the trace is visible, but unequal, and the polarization normal.

Diamond.-I had in possession for a few days some large diamonds which form part of the collection of the Duke of Brunswick, and which had been obligingly lent me by the municipality of Geneva. Unfortunately the season was very unfavourable, and I was only able to avail myself, for the observations, of a few moments during which the sun shone. Besides, the diamonds being cut as brilliants renders the observation very difficult it gives rise to much false light; and on account of the reflection of the facets, one may readily confound the luminous trace with one of its images.

(1) A white diamond of 39 carats gave rise to a trace very visible and nearly white. The greater part of the light emitted by it is due to fluorescence; but I feel certain that another part is due to illumination properly so called; indeed the trace is partially polarized.

(2) Some yellow diamonds of 29, 31, 42, and 80 carats gave rise to a very visible bluish trace, principally due to fluorescence, but also presenting partial polarization*.

*To observe the polarization-phenomena, I operated in the following manner. A pencil of sunlight, concentrated by a quartz lens of long focus, entered the dark room; it was polarized in passing through a Nicol, and then fell upon a screen pierced with a small aperture. A slender pencil

I do not hesitate to attribute these phenomena of polarization to foreign particles. We know, indeed, from Dumas and Stas*, that diamonds leave a spongy residue after combustion; and Sir D. Brewstert recognized with the microscope cavities of various shapes in their interior.

Ice. When very clear and compact, it exhibits very little capacity of illumination. When floating in water in a large glass vessel through which the pencil of rays passes, the trace is very pronounced in the water, but almost invisible in the ice. Cutting a piece of ice into a suitable shape, I had no difficulty in distinguishing in it the trace, which, when viewed through a lens, is seen to resolve itself into minute particles. It is much less visible than that of the purest water I succeeded in obtaining by distillation; its polarization is complete §.

Alum.-A crystal of alum, in which I had faces cut at right angles, gives a trace in the parts where flaws are present, but in other parts does not appear to have any capability of illumination.

Gypsum. I cite from memory a specimen of gypsum from the collection of M. Alph. Favre. I examined it, a long time ago, with the oxyhydrogen light only, and did not recognize in any trace.

it

The whole of these observations appear to me to well establish :

1st. That in non-fluorescent crystals the crystallizable matter by itself is destitute of illuminability.

2nd. That crystals which possess the faculty of being illuminated by returning polarized light laterally, owe that property either to the presence of foreign particles or to defects of crystallization.

of polarized light was thus obtained, which I caused to enter into the diamond through the large front face of the brilliant, and to emerge through the parallel facet (the collet) which generally terminates the rear of a brilliant. The trace was then viewed with a Nicol, through one of the inclined faces that form the posterior pyramid. In these conditions, on turning the Nicol, differences of intensity are observed if the trace is partially polarized.

* Annales de Chimie, 1844, vol. i. p. 15.

† Trans. of Royal Society of Edinburgh, vol. xxiii. p. 1.

For the purification of water by distillation, see Archives, 1870, vol. xxxvii. p. 146.

§ I am here speaking of the naturally formed ice of ponds; there are great differences, according to the specimens. I recall here that Tyndall indicated water proceeding from the melting of ice as the most exempt from particles and illumination-power that he could obtain.