O'Reilly — On Gaseous Products of Great Eruptions. 29 



great depth, and having only the initial angular velocity corre- 

 sponding to that depth, it should lag more and more as it rose to 

 the surface of emission. This lateral pressure and friction would, 

 to a certain extent, explain the comminution of the lava, and the 

 formation of the very fine dust. 



That Krakatoa, and indeed the whole of Java, having for 

 antipod the north-western coast of South America, must in like 

 manner, to some extent, feel the "recoil" of eruptions and earth- 

 quake shocks happening in the latter localities, may be expected, 

 since we have here the occurrence of the exceptional case of land 

 having for antipod land, and as can be seen at a glance of the map 

 exhibited, showing the antipodes of the countries of the Eastern 

 Hemisphere, South America, and part of Asia correspond in a 

 very remarkable manner, while at the same time they represent 

 the most active seats of volcanic and earthquake action. Were 

 the soundings of the ocean and our bathymetrical maps complete 

 in this respect, and could these soundings be verified periodically, 

 it would perhaps be found that, corresponding to the volcanic and 

 earthquake actions which take place in one hemisphere, move- 

 ments of the bottom of the ocean take place in the opposite hemi- 

 sphere, and that thus the deformations resulting from a continual 

 contraction are being balanced, so as to maintain the uniformity 

 of the earth's movement. 



There is one last point relative to the great eruption to which 

 I venture to call attention; it is that of the periodicity of such 

 great outbursts. 



One of the remarks made relative to Krakatoa in the article of 

 Nature already referred to, vol. xxx. p. 10, is, that the volcanoes 

 of the Straits of Suncla had been in a state of quietude during 

 200 years, and that during the latter years a great many earth- 

 quakes took place along the fissure on which they are situated. 

 Now, in the same vol., p. 435, is a very interesting article on the 

 frequency of earthquakes in Japan, It is stated that the Japanese 

 have attempted to prove that earthquakes run in well-defined 

 cycles, a by no means novel or very modern idea. Wernich, in his 

 Geographische Medicinische Studien, says "that severe earthquakes 

 occur in Japan every 20 years." The Japanese journals, working 

 on records relative to the period included between the dates 



