O'Eeilly — On the Dates of Volcanic Eruptiom. 395 



defining precisely the term '■^volcanic eruption'''' " so that a basis 

 may be determined for future scientific investigation." 



As regards the total number above-mentioned, the following were 

 regarded as cases of distinct and separate eruptions : — 



1. xiU normally occurring eruptions of individual volcanoes. 



2. All very markedly active repetitions of these, when they present 



their usual course and character, such as a renewal of lava 

 outflow, formation of new craters after the old ones had 

 become quiescent, etc. 



3. All marked paroxysms of mud volcanoes. 



4. All mud and water outbursts of individual volcanoes. 



5. Ash-showers, more particularly from high volcanoes. 



6. Sudden outbursts of smoke, with explosions, after lengthened 



periods of rest, and during periods of marked earthquake 

 action, when the supposition is allowable, that the lava 

 has not reached the rim of the crater or has opened up for 

 itself subterranean passages. 



7. Markedly strong activity in the cases of volcanoes — such as 



Stromboli, Sangay, Lamangan, etc., which are usually in a 

 state of continual activity, 



8. Sudden elevation of large land surfaces of short duration. 



He then discusses the relative significance and importance of these 

 different classes. On p. 585 he gives two tables of the distribution of 

 eruptions as regards centuries of occurrence and countries. He says, 

 on p. 585, " One of the most striking results furnished by the 

 table is that showing the ■ repetition of volcanic eruptions with 

 reference to seasons." 



" Whilst as regards earthquakes there is generally shown a pre- 

 dominance of winter earthquakes, in the case of volcanic eruptions 

 there is shown a preponderance of these phenomena as regards the 

 summer months, and to such a marked degree that if the sums of the 

 farthest northern eruptions and farthest southern ones be taken 

 together, the result cannot be looked on as purely accidental." 



"It is the same result at which I had arrived for the grouping 

 of earthquakes in general and of earthquake periods, and which seems 

 anew to prove that the great world-shaking earthquakes have their 

 origin in causes other than purely local ones limited to circumscribed 

 territories. 787 volcanic eruptions, for which the data are sufficiently 

 correct, distribute themselves as follows over the different months of 

 the year • the commencement of the eruption being always taken (in 



2 F 2 



