occasion of numerous theoretical extravaganzas, brought us back to the basis of a rational treatment of the subject. Could Karsten not have known of this work in drawing up the papers in the 60th volume of the Annalen?

The explanation which Karsten gives of Moser's images is altogether inadmissible and may be easily refuted. He thinks that, because similar images can be produced by the aid of electricity, Moser's images must be of electrical origin. He thinks that "if two bodies, differing in any respect from each other, come in contact, an electrical current is produced!" and that this is the cause of Moser's images.

The generation of an electrical current by the contact of two heterogeneous bodies, which Karsten seems to intimate in this passage, will not be granted by the most zealous of the adherents of the contact theory; but granting even the existence of such a current, it could not produce any image, as the researches of Riess prove.

That electrical tension alone, without repeated discharges between the body and the plate, is not sufficient to produce electrical images has been shown by Know in a paper "On electrical figures and thermography," (Pog. Ann., LXI, 569,) in which he has proved the untenableness of Karsten's view as to the electrical origin of Moser's images.

The rest of the contents of Know's memoir will be mentioned subsequently in the proper place.

§ 77. Electrical breath images.-Riess placed a metal stamp on a shining pitch surface, and upon the stamp a small metal weight connected by a silver wire with the knob of the spark micrometer, receiving electricity directly from the conductor of the machine, while the other knob of the spark micrometer, one-half line from the first, was in conducting connexion with the ground.

The machine being now turned, electricity accumulates upon the first knob of the micrometer and upon the stamp, until a discharge takes place by the passage of a spark between the two knobs; continued turning will charge and discharge the stamp anew. The discharges follow more rapidly the closer the knobs of the spark micrometer are together.

After several revolutions of the machine the stamp may be removed, the plate breathed upon, when a shining image of the stamp shows itself on the dull ground.

It is indifferent for the success of this experiment which electricity is used.

Such images may also be produced on glass and mica, but on these substances they are often imperfect.

The simple breath image, Riess says, in caused by repeated electrical discharges taking place in opposite directions between the model and the insulating plate. The electricity communicated to the model passes over to the plate, then back to the model, when the latter is discharged by the spark micrometer; thus a motion of the same kind of electricity arises, first downward and thus upward. Since the discharges between a bad and a good conductor are never perfect, electricity, both of the kind used and the opposite kind, remain upon the

insulating plates, which are therefore in the condition to produce dust figures, often even dust images.

By simply electrifying the stamp, the arrangement being the same as for producing dust images, no breath image appears. The alternate charge and discharge of the stamp are essentially necessary for the formation of these images.

By laying a plate of mica on a pitch plate, and placing a metal stamp on this, a double discharge of the same kind of electricity takes place in the same direction in electrifying the stamp, namely, from the stamp to the upper surface of the mica, and from the under surface of the mica to the pitch plate. When a spark is communicated to the stamp from a positively charged jar, the pitch surface, when dusted, shows a yellow image of the stamp, surrounded by positive dust figures. If, therefore, in this arrangement of the stamp alternate charges and discharges are brought about, the conditions for forming manifold breath images are fulfilled.

A pitch surface being covered with a mica plate and a stamp placed on it, the latter was charged and discharged by the spark micrometer. After twenty revolutions the upper surface of the mica showed a perfect breath image, but the under surfaces and that of the pitch presented a most imperfect one.

These images are so frequently imperfect because pitch and mica adhere closely together in consequence of the electricity remaining after each discharge and subsequent discharges is conveyed to places which lie scattered beyond the image surface; but a metallic plate being substituted for the pitch plate, a perfect breath image is obtained on the upper and lower surfaces of the mica and on the metallic surface.

The visibility of the breath images is to be explained, according to Riess, by the fact that the surfaces are freed by electrical discharges from the film of foreign matter with which they are generally covered; and he has even proved such a cleansing of the surface by images on metal. On a perfectly insulating mica surface Riess produced a breath image, and the place where the image appeared conducted as well as a fresh surface of mica, thus showing that it had been freed from the stratum covering this spot.

In most cases breath images are produced by such a cleansing action, but they can be excited also by soiling the plate.

On a fresh mica surface an obscure image of a stamp was obtained on a shining ground. On an old surface, which electrified by forty revolutions, gave a bright breath image; one hundred revolutions produced a dull image.

The various kinds of dull breath images depend upon the condition of the plate used and of the stamp, and also upon the strength of the electricity; the clear images appear more frequently only because soiled plates and the least possible electricity are generally used.

The origin of the breath images, like that of the breath figures, is to be ascribed to a change which the electrical discharge produces in the stratum covering the plate, and consists in an increase or diminution of this stratum, according to circumstances.

A spark thrown upon a metallic surface injures it when perfectly

clean, but leaves it unchanged if it is soiled or tarnished. This is the case, in forming breath images on metals. A very small number of discharges having passed between a metallic surface and one of m ca covering it, the intermitting discharge begins in the foreign stratum on the surface of the metal, and the metal remains uninjured; but when the stratum is destroyed, and the breath image is produced, and the discharges are continued, the latter then begin on the metal itself, which is thus changed. Such images, appearing without breathing, and representing some parts of the stamp in brownish colors, Riess produced on silver with from fifty to sixty revolutions.

§ 78. Electrolytic images.-If the blunt point of a platinum needle be placed on a paper moistened with a solution of iodide of potassium, and lying on a metallic plate connected with the ground, a brown spot will appear under the point if the needle is electrified positively, but there will be no spot if it be negatively electrified. Using positive and negative electricity one after the other in any order, the coloring remains even when the quantity of negative electricity far exceeds that of the positive.

This fact explains the electrolytic images, which Riess has invented for proving the correctness of the view presented above, on the formation of breath images by alternating discharges.

A piece of card paper, moistened on one surface with a solution of iodide of potassium, was laid on a metallic plate connected with the ground, and then covered with a plate of mica. A stamp was placed on the mica, and, being loaded with a weight of 2 to 14 ounces, was connected with the spark micrometer, whose knobs were a line asunder. After twenty revolutions of the machine, positive electricity continuing to pass between the knobs, a very sharp image appeared on the paper in which the letters of the stamp appeared with a brown color.

The explanation of this phenomenon, according to the above, is easy. As in breath images, the stamp being charged with positive electricity, it passes from the lower surface of the mica to the metal plate, and thence through the moist paper; by this passage of the E to the metal plate the iodide of potassium is decomposed; as soon as a discharge takes place between the knobs of the spark micrometer, an opposite current sets in between the metal plate and the mica; the +E now returns to the mica, and the E through the moist disk to the metal. While the E goes to the metal the iodide of potassium is decomposed, and this effect is not destroyed by the discharge in the opposite direction.

It is to be remarked that the passage of the + E from the mica to the metal takes place gradually, while the discharge in the opposite direction happens instantaneously.

The same experiment being repeated in the same manner with — E, no image is obtained, but only irregular brown spots.

This also may be easily explained; the negative electricity goes gradually to the metallic plate, while the passage in the opposite direction is instantaneous; thus, a greater quantity of positive electricity returns at once to the metal plate, and passes more readily to such points as lie beyond the image surface.

To obtain an image with negative electricity, care has only to be

taken that the quantity of + E which returns on the discharge between the knobs to the metal plate, shall be less, which is attained by bringing the knobs of the spark micrometer closer together.

SECTION FOURTH.

ELECTRICAL SPARK AND BRUSH.

§ 79. Faraday's researches on the spark and brush.-Without going into the theoretical disquisition, mentioned in another place,* which Faraday has given upon the spark and brush, I will present here only the most important facts which he has obtained in his experiments upon these phenomena of light.-(Pog. Ann., XLVII and XLVIII.)

In order to compare the resistance which different gases presented to the passage of sparks, with the corresponding resistance of the air, Faraday used an apparatus, a sketch

of which is represented in fig. 72. Two small knobs, s and S, connected with the conductor of an electrical machine, were placed opposite to two larger knobs, and L, in conducting connexion with the ground. The diameter of the balls was as follows:

Fig. 72.

The constant interval v between 8 and I was 0.62 of an inch; the interval u between S and L was variable.

It would have been better if the two small balls s and S had been perfectly equal in size, and 7 and L also equal; much more reliable conclusions could then have been drawn from these experiments.

The two balls 8 and 7 were placed in a receiver, which could be exhausted and then filled with different gases.

The receiver being filled with air under the pressure of the atmosphere, the sparks passed alternately at u and v, when the intervals at u were between 0.6 and 0.79 inches. When the interval at u was less than 0.6 the sparks always passed here, but if it was greater than 0.79 the sparks then always passed at v.

* See § 24 in the Report for 1856.

Similar results were obtained when other gases were in the receiver under the atmospheric pressure. There were two limits for the interval at u, between which the spark passed at one time at u, at another at v; the interval at u being less than the least of these limiting numbers, the spark passed always at u, but being greater than the greatest of these numbers it always took place at v. The following table indicates the limits at u for different gases, v having the constant value of 0.62 inch :

which does not coincide very well with the former results, a proof that these numbers do not afford sufficient grounds for forming a conclusion.

That within certain limits of distance at u the spark takes place alternately at u or v, and consequently that there is not a single permanent value of u for each gas, over which the spark always happens at v, but under always at u, depends upon accidents (such as particles of dust floating in the air) of which we can give no account.

If at one of the intervals a spark once passed there was generally s strong tendency in it to appear at the same interval again.

It is a remarkable circumstance that the range of distance u should be much less when s and S are negative than when these balls are