Progress of Invention, 235 



tures of which are placed opposite to, and at a small distance from, 

 each other. To dilute the magnesium powder, and thus render the 

 feed more gradual, and the light more diffused, it is mixed with 

 fine sand before being placed in the reservoir. And to prevent the 

 tubes through which the mixture of sand and magnesium issues 

 from being choked with magnesia, on account of combustion of the 

 magnesium taking place within them, coal gas is transmitted through 

 them with the mixture, and burns along with the magnesia. 

 Finally, to remove the small quantity of magnesia which, forming 

 at the aperture of the tubes, might ultimately stop them up, the 

 sand, after separating from the magnesium during the combustion 

 of the latter, falls into a small receiver ; and this, tilting over every 

 few minutes, on account of the weight of the sand, gives motion to 

 a wire to which are attached nippers that seize the magnesia and 

 separate it from the tubes. When the sand has fallen out of the 

 receiver, it again assumes its ordinary position. The cost of this 

 lamp is, however, very considerable. The amount of coal-gas used 

 with it is, indeed, very trifling, but the magnesium is very expensive, 

 and will probably continue so. This very serious objection is likely 

 to be got rid of by the use of the oxide of magnesium, which M. 

 Carlevaris, of Geneva, has found to answer, in conjunction with the 

 oxy-hydrogen flame, as well as the metal itself. Oxide of magne- 

 sium is in this case volatilized, though to a very small extent ; but 

 this minute quantity of vapour imparts a pure white light that may 

 be used either for ordinary illumination or for photography, having 

 all the properties of solar light. The effect is due to oxide of 

 spongy magnesium, and this may be conveniently produced from seve- 

 ral of the magnesium salts, especially from chloride of magnesium or 

 carbonate of magnesia. For this purpose a small piece of chloride, 

 for example, is placed on a prism of gas-retort graphite, and the 

 flame of the burning gases is thrown upon it. The use of pure 

 oxygen is not indispensable ; common air and hydrogen answer 

 extremely well. Also, the chloride, after having been compressed 

 into the form of a prism or cylinder, may be merely introduced into 

 the flame, no graphite being employed. 



Prevention op the Spark produced by the extra Current. — 

 In electro-magnetic experiments, the oxidation caused by the 

 sparks due to extra currents, is highly inconvenient, as it prevents 

 perfect metallic contact, and thus greatly diminishes the electro- 

 magnetic effect. There are two kinds of extra current. One of 

 them is opposite in direction to the battery current, and is produced 

 when connection with the battery is established. It is perceptible 

 chiefly by its neutralization to a greater or less extent of the battery 

 current. The other, which is in the same direction as the battery 

 current, is produced when connection with the battery is inter- 

 rupted. It retards the demagnetization of the electro-magnet, and 

 causes oxidation so as to prevent perfect metallic contact between 

 the different portions of the conductor. The latter evil is pre- 

 vented by a very simple means, due to Dr. Dugardin, of Lille. He 

 provides a separate conductor for the extra current. For this 

 purpose he permanently connects the poles of the electro-magnet 



