powerful "sensitizer" (see the above work, 1873, p. 88); that is to say, it augments their sensibility in consequence of its fixing chemically the iodine or bromine as it becomes liberated during exposure. That this action occurs chiefly in the blue finds its solution without doubt in this, namely that the blue rays are more eagerly absorbed by the wet film than are the others.

As has been already observed, the sensibility of the dry bromide of silver diminishes gradually from the blue to the red. With bromide-of-silver plates as prepared by me I could see nothing of the phenomenon which I had noticed as occurring in such a marked manner with the English bromide-of-silver plates spoken of above, namely a falling off of the sensibility from the violet to the blue, and an increase thereof from the blue to the green. The explanation above offered as to the action of nitrate of silver upon bromide of silver induced me, however, to conjecture that the English bromide-of-silver plates must contain some substance that absorbs the green in a greater measure than the blue. It is not unusual to give dry plates a coating of substances of the most varied kinds, such as gallic acid, caffeine, or morphine, all which bodies fix iodine and bromine and exercise a sensitizing action; occasionally, too, a yellowish colouringmatter is added thereto, with the view of retarding the "chemical" blue light thereby. The optical demeanour of these "preservatives" may be looked upon as a matter not yet in any way understood; neither, indeed, is their favourable influence placed beyond all question.

The plates of Wortley contain nitrate of uranium, gum, gallic acid, and a yellow colouring-matter as a coating. In order to ascertain whether this coating exercised any action, I washed one of these plates with alcohol and water, and obtained in fact by so doing a plate that no longer gave any indication of an augmented sensibility in the green. I now made an attempt to impregnate bromide of silver with a substance that absorbs especially the yellow rays and fixes free iodine or bromine, in the hope of thereby heightening the sensibility for yellow. I selected coralline, which Professor Liebermann most kindly placed at my disposal. A solution thereof, when greatly diluted, gives in the spectroscope an absorption-band between D and E; in a more concentrated solution the absorption becomes widened out until it reaches beyond D, while, on the other hand, the blue near F is transmitted to a pretty considerable amount.

I dissolved coralline in alcohol and added some of it to my bromide collodion, so that it became of a strong red colour. Dry bromide of silver plates prepared with this collodion were of a decided red tone. On exposure to the spectrum they bore out my anticipation; that is to say, the plates proved sensitive

in the indigo, from there their sensibility decreased till the light blue, became weak at F, then increased again, and in the yellow was found to be almost as efficient as in the indigo. Thus a method was attained for preparing bromide-of-silver plates that are acted upon almost as strongly by a colour hitherto held to be chemically inert, namely yellow, as they are by indigo, which hitherto has been held to be the colour possessing the strongest chemical energy.

After this experiment I was justified in hoping that some other bromide-fixing substance endowed with a powerful absorption of the red would in like manner heighten the sensibility of bromide of silver for red. Such a body I met with among the green aniline products. The body in question absorbed in a marked manner the red rays in the middle between D and C; when further concentrated this absorption became extended towards D, while the yellow, green, and blue were transmitted almost unimpaired. In point of fact a collodion coloured with this green proved sensitive to light as far as into the red.

The sensibility fell off from the indigo to the yellow, then increased again; and at the identical place where the above-mentioned absorption-band was visible a powerful action in the red was manifested.

From these experiments I think I am pretty well justified in inferring that we are in a position to render bromide of silver sensitive for any colour we choose that is to say, to heighten for particular colours the sensibility it was originally endowed with. To effect this, all that is required is to add thereto some substance which promotes the chemical decomposition of the bromide of silver and absorbs the particular colour in question but not the others. Perhaps we may even arrive at this, namely photographing the ultra-red as we have already photographed the ultra-violet. The photographic inefficiency of certain colours, which has hitherto proved so great an obstacle, would in that case be surmounted. The following experiment shows to what extent this is practically substantiated. A photograph was taken of a band of blue upon a yellow ground. Employing an ordinary collodion plate with iodide of silver, I obtained thereupon a white band upon a black ground. On a bromide-of-silver and collodion plate, where the action of the blue and the yellow is equally powerful, it was to be anticipated that no effect was to be obtained. I therefore placed in front of the object-glass a disk of yellow glass, which absorbed the blue light and transmitted the yellow unimpaired; and then, after an exposition of suitable duration, I obtained in point of fact a dark band on a light ground.

The matter is not alone of practical but also of scientific in

terest. Hitherto it has been held that the haloid salts of silver were modified chemically only by the rays which they absorb to a marked extent (Schultz-Sellack, 'Reports of the German Chemical Society,' 1871, p. 211); and, moreover, the influence of "sensitizers" has been partially called in question (SchultzSellack, 'Photographic Communications,' seventh annual publication, p. 301).

My investigations show that, with respect to the sensibility of photographic plates to the action of light, not only does the optical aptitude for absorption on the part of the sensitive silver salts themselves, but also the optical aptitude for absorption on the part of the substances mixed therewith play a prominent part.

Further experiments on this point are in progress.

Berlin, October 1873.

XXXIII. A Theory of the Effects produced by Fog and Vapour in the Atmosphere on the Intensity of Sound. By Professor CHALLIS, M.A., F.R.S., F.R.A.S.*

THE

HE experimental results obtained by Dr. Tyndall relative to the intensity of sound under various conditions of the atmosphere (as published in the Proceedings of the Royal Society, No. 149, pp. 58-68) are of so much scientific importance that it seemed to me desirable to endeavour to account for them theoretically. The questions they raise are clearly hydrodynamical; and I think I shall be able to show that the appropriate answers may substantially be given by reference to propositions in hydrodynamics which I have previously discussed in the pages of this Journal.

It is evident from the experiments that the phenomena to be explained are of two kinds-those resulting from Fog, or the presence in the air of vapour in a visible form, and those due to the admixture with the air of invisible vapour. As to Fog, or Haze, I shall at once assume that it is to be attributed to the suspension in the air of extremely minute globules of water. This supposition is justified by phenomena of the Fog-bow, which indicates by its form an action of spherical drops of water on light the same as that which takes place in the case of the primary rainbow; and its freedom from colour gives evidence of the extremely small size, as compared with that of rain-drops, of the suspended globules.

Let us conceive aërial vibrations capable of producing sound to be generated and transmitted in air loaded with an immense

* Communicated by the Author.

number of these small globules. The problem of determining the velocity and condensation of the air due to the reaction of a single small sphere on which such vibrations are incident, I have already discussed and applied on several occasions. (See the Number of the Philosophical Magazine for June 1864, pp. 457 & 458, that for October 1865, pp. 262 & 263, and the solution of Example VI. in pp. 279–287 of 'The Principles of Mathematics and Physics.') Suppose the velocity at any point due to the reaction of the sphere to be resolved along, and perpendicularly to, the radius vector drawn to the point from the sphere's centre, and let the former resolved part be U, and be reckoned positive in the direction from the centre, and the other be W, and be reckoned positive when it is in the same direction as that of the incidence; also let σ be the condensation at the same time at the same point. Then if r and be polar coordinates of the given point, the origin being at the centre of the sphere and being the angle which the radius vector makes with an axis drawn in the direction of incidence from the origin, and if b be the radius of the sphere, and T the velocity of the incident vibration at the time t in a plane through the origin transverse to the axis, the solution of the above-mentioned problem gives

Now any reflection of the incident waves, caused by the reaction of the sphere, will be proportional to ao, and must be extremely

small on account of the smallness of the factor

b dT

a dt'

If, for

and

the waves, the maximum value of that factor becomes

b

2πmb

λ

therefore, by reason of the ratio, is excessively small for all such values of m and λ as those which are applicable to the experiments.

b2

2r2

Again, on account of the small size of the globule, the factor will become excessively small at small distances from its centre, because such distances may still be many multiples of b. Thus, by reason of the two factors combined, the reflection from a single globule may be so extremely minute at all sensible distances from its centre, that the sum of the reflections from a very large number of globules contained within a given finite space might only generate reflex waves of moderate magnitude and coming

from a great depth. The experiments of Dr. Tyndall showed that this is actually the case in a foggy or hazy atmosphere; and accordingly this fact, so contrary to previous expectation, may be considered to be accounted for by the present theory.

Taking now into consideration the velocities U and W, it may be proved (as is shown in the discussions above cited) that these velocities are such that just as much fluid crosses any plane perpendicular to the direction of the propagation of the incident waves as would have crossed if the globule had been absent. As this is true with respect to each globule, it follows from the principle of the coexistence of steady motions, that the mean effect of the reaction of all the globules is simply to augment the velocity in the incident waves by contracting the space in which the fluid moves, just as the rate of the flowing of a river is increased by contraction of its channel. By this means an impetus would be given to the sound-waves; and it might thus happen that, in the absence of countervailing causes, sound would be audible in a hazy atmosphere at unusually great distances. This in fact appears to have occurred in the experiments of July 1 and October 30.

During a fall of rain or snow, the drops or the flakes might by their size produce a sensible amount of reflection, and so diminish the intensity of the sound; but, on the other hand, by their occupation of space they would, in the manner explained above in relation to the globules of vapour, give impulses to the vibrations; so that upon the whole (as appears from experiment to be the case) rain or snow falling might have little effect on the intensity of sounds.

It is to be observed that the theory points to no change of the rate of propagation of sounds by the action of visible vapour, but exclusively to an alteration of their intensity.

Let us now suppose vapour in an invisible state to be mixed with the air. The vapour is connected with the air only by a kind of mechanical suspension; yet there is reason from experiment to conclude that the elasticity of the compound differs very little from that of the air itself. So long, therefore, as the mixture is not irregular, there would appear to be no reason for the production of reflected sound. But, to use Dr. Tyndall's terms, an irregular or "flocculent" admixture would render the air "acoustically opaque." This effect may be conceived to take place as follows.

If it were possible that two portions of an elastic fluid of different densities and the same elasticity could be in juxtaposition on opposite sides of the same plane, in any case of the propagation of waves from one portion to the other reflection would take place at the plane of separation, whether