A new explosive has been discovered by M. Roca, a French engineer, who communicates an account of it to _Le Génie Civil_. The discovery was due entirely to scientific induction from some experiments made upon different specimens of dynamite, with a view to the determination of the effect on the explosive force of the various inert or at least slowly combustible substances with which nitro-glycerine is mixed to produce the dynamite of commerce. Of late, in place of the infusorial earth which formed the solid portion of Nobel's dynamite, such substances as sawdust, powdered bark, and even gunpowder, have been used, probably for the sake of economy alone, without, except in the latter case, any reference to the influence which they might have upon the combustion of the nitro-glycerine; but M. Roca, in testing a variety of samples, was struck by the difference among them in regard to energy of explosion, and discovered that if a portion of free carbon, sufficient to combine with the oxygen disengaged from the nitro-glycerine, was present at the moment of detonation, the effect was greater than where, as in the case of gunpowder, the solid portion alone furnished oxygen enough to burn all the free carbon, without calling upon the nitro-glycerine for any. In fact, it appeared from experiment that the dose of carbon might with advantage be so great as not only to be itself oxidized into carbonic oxide by the oxygen of the nitro-glycerine, but to reduce the carbonic acid developed by the explosion of the latter itself into carbonic oxide. The limit of the advantageous effect of free carbon ceased here, and if more were added to the mixture, the cavities formed by the explosion in the lead cubes used for test were found simply lined with soot; but up to the limit necessary for converting all the carbon in the dynamite into carbonic oxide, the addition of a reducing agent was shown to be an important gain. This was confirmed by theory, which shows that pure nitro-glycerine, which is composed of six parts of carbon and two of hydrogen, combined with three times as much nitric acid and water, decomposes on explosion into six parts of carbonic acid, five of watery vapor, one of oxygen, and three of nitrogen, while the addition of seven more parts of free carbon to the mixture causes the development, by explosion, of thirteen volumes of carbonic oxide, five parts of watery vapor, and three of nitrogen, or twenty-one volumes of gas in place of fifteen. As the power of an explosive depends principally on the amount of gas which results from its sudden combustion, it was evident that the addition of pure or nearly pure carbon, in a condition to be readily combined with the other elements, ought to increase materially the force of nitro-glycerine, and M. Roca experimented accordingly with an admixture of sugar, as a highly carbonized body immediately available, and found that three parts of this, mixed with seven parts of