principal heights. The Delaware, Susquehanna, and Potomac rise west of the Alleghany Front, which they cross, and, continuing eastward, traverse the Alleghany ridges and the Blue Ridge to reach the Atlantic. From among the Alleghany ridges of Virginia the James and Roanoke flow through the Blue Ridge eastward. New River, on the contrary, has its source east of the Blue Ridge in North Carolina, and runs northwest across the Blue Ridge, the Alleghany ridges, and the Alleghany Front, to the Ohio. It is thus a general fact that the streams of the Appalachian ranges are not controlled by the mountains. The ridges pursue their courses, and the streams, passing across the ridges, pursue independent courses. The discordance is one of the most marked features in the topography, and it gives rise to many picturesque water gaps. It is due to the fact that the transverse river channels are older than the valley ridges. Within the valley the brooks and creeks have arranged themselves usually in systems of pairs. Flowing southwest, a brook meets its fellow running northeast, and together they turn southeast or northwest to traverse a ridge. In the valley beyond the ridge they are joined by a pair similar to their own courses before their union. Beyond a second ridge or a third, the growing creek may for a time flowing northeast or southwest, but it will presently pass out by another water gap. Ultimately it falls into one of the great transverse rivers. This arrangement of parallel brooks which swell the volume of a creek generally flowing at right angles to their courses, resembles a vine from whose central system branches are trained on a trellis. Although most conspicuously developed in the Alleghanies, this trellis system of drainage is common in regions where beds of hard rock lie steeply inclined to the general surface. The parallel branches of the system are controlled by the parallel ridges between each two pairs. Thus it appears that the hard rocks have to this extent influenced the arrangement of the streams.
Petroleum-Lamp Accidents.—The recent report of Mr. Alfred Spencer, an officer of the control department of the London County Council, on petroleum lamp accidents, and the measures necessary for preventing them, is a very important and practical document. His conclusions regarding their safe construction and proper management are as follows: (1) The oil reservoir should be of strong metal, properly folded and soldered at the joint, and should not be of china, glass, or other fragile material. (2) There should be no opening between the reservoir and the burner, other than through the tube which holds the wick, and this tube should be extended to within a quarter of an inch of the bottom of the reservoir, and should have no opening into the reservoir except at its base. (3) The burner should be securely attached to the reservoir, preferably by means of a strong and well-made screw attachment. (4) There should be no openings through which oil could flow from the reservoir should the lamp upset. (5) Every table lamp should have a broad and heavy base, to which the reservoir should be strongly attached. (6) Wicks should be soft and not tightly plaited, and should quite fill the wick-tube without having to be squeezed into it. (7) Wicks should be frequently renewed, and before being put into lamps should be dried at a fire and then immediately soaked with oil. (8) The reservoir should be filled with oil before the lamp is lit. (9) The lamp should be kept thoroughly clean, all oil should be carefully wiped off, and all charred wick and dirt removed, before lighting. (10) When first lit, the wick should be partially turned down, and then gradually raised, (11) The wick should not be left turned down, as there is then a greater liability to explosion in lamps of unsafe construction. (12) Lamps which have no extinguishing apparatus should be put out as follows: The wick should be turned down until there is only a small, flickering flame, and a sharp puff of breath should then be sent across the top of the chimney, but not down it. (13) Cans or bottles used for oil should be free from water and dirt, and should be kept thoroughly closed.
The Serum Treatment of Disease.—It is stated in the British Medical Journal that the serum treatment of disease probably originated in the observation made by Von Fodor, in 1887, that blood when drawn from the body had a distinct bactericidal action. "Nuttall and others then pointed out that