rocks disintegrated in this way merely fall to pieces, the hard portion remaining in the shape of sand or pebbles. The transportation of this by streams produces what the author calls corrasion. In this way the bed of a stream is widened and deepened, but the work is also facilitated by the ceaseless action of water in dissolving the rocks.
The mechanical wear or erosion by a stream depends largely on its velocity. "A stream of water flowing down its bed expends an amount of energy that may be measured by the quantity of water and the vertical distance through which it descends."
The velocity of a stream would continually increase if none of its energy was consumed in friction, but very much of it is so consumed, and reappears in innumerable forms of movement or subsidiary currents. It is by some of these that the work of transportation and erosion is largely done. Bat a stream may be overloaded with detritus, and its corrading power correspondingly diminished. "Only with a partial load does a stream wear its bottom." Of the Colorado plateau the author says that the erosion which began with the first lifting of a part above the ocean has progressed continually to the present time. The total uplift has been about 12,000 feet; only 7,000 feet remains, that being the present altitude above the level of the sea. Five thousand feet of the general surface has been removed, and an amount greater by several thousand feet has been corraded by the rivers.
Improved Railway-Signal.—A simple and effective railroad-signal, in use on the Boston, Lowell & Nashua Railroad, is described in the Scientific American. A single-cell Callaud battery is connected to the two rails at one end of a given section of the line—say, two miles in length—each section being insulated from adjoining sections. At the other end the signal has an electro-magnet similarly connected to the two rails. When the circuit is closed, as is normally the case, the magnet is excited and the signal controlled thereby so as to show that the line is clear. But, when a train runs on the section, then a shorter circuit is made by the wheels and axles, and the current returns to the battery by this course, instead of passing through the signals. The magnet ceases to attract, and the signal by mechanical means is at once turned, to indicate danger. It is obvious that this must occur as long as a single car remains on the track, or when the circuit is broken by a displaced or ruptured rail or any other cause. Hence the device may be applied over an entire line, and will indicate the condition of every section to a train about to enter on the same. It is found to be operative in all weathers.
Powder-Paper.—A substitute for gunpowder has been invented in England, called "powder-paper," viz., paper impregnated with a mixture of potassic chlorate, nitrate, prussiate, and chromate, powdered wood-charcoal, and a little starch. The powder-paper is rolled into the shape of a cartridge of any required length or diameter. The manufacture involves no danger, it is said; no explosion can take place except by way of contact with fire. The powder-paper leaves no greasy residue on the inside of the gun; it also produces less smoke, gives a less violent recoil, and is less impaired by humidity than gunpowder. With equal charges, by weight, of gunpowder and powder-paper, the penetrating power of the latter is 16 greater than that of the former.
October Meteor-Shower.—In a letter to the Tribune, dated October 19th, Mr. Daniel Kirkwood states that shooting-stars in unusual abundance were observed by several trustworthy witnesses at Bloomington, Indiana, on the evening of the 18th, from six hours forty-five minutes to nine hours. The meteors appeared to radiate from Auriga, or rather from a point between Taurus and Auriga. Most of the meteors were small, though two of them possessed extraordinary brilliancy. In a small work on comets and meteors, published three years ago, Mr. Kirkwood called attention to the fact that meteoric showers had been observed at the same period of the year in 1436, 1439, 1743, and 1798. Returns of the shower were observed in 1838 and 1841. He recommends that a careful watch be kept in future about the same period—say, from the 16th to the 20th of October.