examine, and which have been formed in his investigations of the perforation of rocks by gaseous explosions. In these experiments, channels were opened through the granite by the gases of nitroglycerin, showing on their vitrified surfaces all the stages, from the drawing out of a thin pellicle of melted glass to the formation of a perfect spherule. Grains of two different categories may be distinguished with the microscope in the dusts produced in the trituration of rocks by the violent passage of gaseous explosions. Some can not be distinguished from those produced by simple mechanical pulverization. But, besides these materials—angular grains of quartz, feldspar, and mica—we find abundance of perfect, or nearly perfect, spheres, opaque and black or slightly translucent and brownish, with shining surface and often the characteristic little neck. Identical elements are found in the dust derived from different rocks that have been submitted to experiment, but with features that vary in each of them.
Dust obtained from the lava of Vesuvius exhibits the globuli-form character in the highest degree. Nearly all the matter is in the condition of black globules of various dimensions, but always very light, and sometimes having a tubulure. The abundance of these globules is manifestly associated with the relatively easy fusibility of the rock, which is also represented in the constitution of the general glaze with which all the parts that have been in contact with the incandescent gases are covered. The identity of these globules with those that exist so abundantly in the atmospheric dusts and marine sediments can not be contested. Till now, says M. Daubrée,[1] the general opinion, and the only one possible, has been to connect the origin of these globules with the arrival of cosmic masses in the atmosphere; and we may add now to the arguments already presented in support of this thesis, the results afforded by the gaseous trituration of meteoric rocks. The dust furnished by a stone cylinder that fell from the sky in 1888, at Pultusk, was marked by innumerable globules, associated with fragments of peridote and entastite, and with metallic granules which had preserved their ramified form and their adherence to lithoid minerals. What we have said shows also that terrestrial rocks, as well as meteorites, may engender these globules. Hence the arrival of meteorites into the atmosphere incontestably contributes greatly to the production of the brilliant globules so abundant in aërian and aqueous sediments. It seems also to be established that the openings of diatremes[2] have an active part in
