This granted, let us see how the apparatus works:  In measure as the water rises in the reservoir, the siphon gradually loses weight, and its extremity, B D H, is finally lifted by the thrust, so that the entire affair revolves upon the studs, K, until the chain becomes taut.  The apparatus then ceases to rise; but the water, ever continuing to rise, finally reaches the apex, b, of the smaller siphon, and, through it, enters the air chamber and fills it.  The equilibrium being thus broken, the siphon descends to the bottom, becomes primed, and empties the reservoir.  When the level of the water, in descending, is at the height of the small siphon, a b c, this latter, which is also primed, empties the chamber, F, in turn, so that, at the moment the large siphon loses its priming, the entire apparatus is in the same state that it was at first.

In short, when the water enters the reservoir, the siphon, movable upon its base, rises to the height at which it is desired that the flow shall take place.  Being arrested at this point by the chain, it becomes primed, and sinks, and the water escapes.  When the water is exhausted, the siphon rises anew in order to again sink; and this goes on as long as the period of irrigation lasts.

This apparatus, which is simple in its operation, and not very costly, is being employed with success for irrigating several meadows in the upper basin of the Allier.—­Le Genie Civil.

* * * * *

ASSAY OF EARTHENWARE GLAZE.

Lead oxide melted or incompletely vitrified is still in common use in the manufacture of inferior earthenware, and sometimes leads to serious results.  To detect lead in a glaze, M. Herbelin moistens a slip of white linen or cotton, free from starch, with nitric acid at 10 per cent. and rubs it for ten to fifteen seconds on the side of the utensil under examination, and then deposits a drop of a solution of potassium iodide, at 5 per cent. on the part which has been in contact.  A lead glaze simply fused gives a very highly colored yellow spot of potassium iodide; a lead glaze incompletely vitrified gives spots the more decided, the less perfect the vitrification; and a glaze of good quality gives no sensible color at all.—­M.  Herbelin.

* * * * *

ON THE ELECTRICAL FURNACE AND THE REDUCTION OF THE OXIDES OF BORON, SILICON, ALUMINUM, AND OTHER METALS BY CARBON.[1]

   [Footnote 1:  Read at the recent meeting of the American
   Association, Ann Arbor, Mich.]

By EUGENE H. COWLES, ALFRED H. COWLES, AND CHARLES F. MABERY.