October 13, 1923] 



NATURE 



547 



formed blisters, some of which, becoming soUdified 

 and remaining hollow, finally crashed in, owing to the 

 lateral fissures'. The emission of lava continued slowly 

 until July 18, when the fiery torrent appeared to have 

 solidified in the crater-mouth. The area covered by 

 the lava is about three square kilometres, as estimated 

 from the photographs taken by me from the hydro- 

 plane M. 28, kindly placed at i^y disposal by Signor 

 Mussolini. 



From the phenomena observed during the eruption, 

 it can be seen that its progress was in direct relation 

 to the mass of the lava emitted, and the various in- 

 cidents were the consequences of special local con- 

 editions. If the structure of Etna were homogeneous, 

 that is to sav without hollows or fissures, the molten 



of the eruption of Etna. The hypothesis of radial 

 fractures which split the volcano at its base is not in 

 harmony with the observed phenomena, and is contrary 

 to the principles of the statics of liquids. 



In this eruption it has been observed that the 

 explosions were due to the detonation of explosive 

 mixtures of volcanic gases — hydrogen, carbonic oxide, 

 and methane — which are given off by the lava, and, 

 when collected in subterranean cavities, form ex- 

 plosive mixtures with the oxygen of the air. The 

 explosions were strongest in those parts of the fissures 

 where deep chambers had formed in which the gases 

 could collect, while towards the uncovered portions 

 of the lava-canal there were milder explosions, with 

 only small jets of lava. Later, when along this canal 



Fig. 2. — The north-east crater at the beginning of the eruption of May 1923. 



\_Photo : G. Ponte. 



lava would not have departed from its principal 

 eruptive conduit, and the eruption would have de- 

 veloped in the central crater. The passages which 

 abound in the lava-flows on the slope of the volcano 

 represent, however, so many subterranean routes which 

 the molten lava could follow through a breach in the 

 principal conduit, which might be formed by the simple 

 collapse of weak parts of its walls or by breaking 

 through where the rock was corroded by acid vapours. 

 We do not know the changes that may have taken 

 place along the epi-subterranean canal during the 

 present eruption, but if its main vent near the principal 

 eruptive conduit is still open, we can assume that with 

 any renewed rise of the magma the lava will follow the 

 same route. If, on the other hand, the breaches in 

 the main pipe have been closed, the magma will re- 

 appear at the central crater until other subterranean 

 routes are opened. 



It is not possible to give a more explicit explanation 



NO. 2815, VOL. 112] 



small cones were formed with corresponding explosion- 

 chambers, the noises became intense. At the mouths 

 of some of these small explosion-cones, there were often 

 seen hissing darts of flame like those of powerful 

 oxyhydrogen jets. These flames, due to the burning 

 of the volcanic gases, have been observed at other 

 volcanoes. 



Various experiments were made during this important 

 eruption. Of particular interest were the successful 

 attempts to reduce, or even to stop for a short time, 

 the explosions at some of the craters near their mouths 

 by introducing carbon dioxide gas, which prevented 

 the combustible gases from meeting with the oxygen 

 of the air. In another experiment, nitrogen was blown 

 through the liquid lava in order to carry away the gases 

 given off, and to enable them to be collected without 

 contamination by the air. This was carried out by 

 means of a special apparatus, already described in the 

 Rendiconti della Reale Accademia dei Lined, vol. xxxi.. 



