Tlie Activo Volcano on Tana, New Hebrides. 81 



the outbursts of Stromboli are regulated by atmospheric 

 conditions has any foundation, is still open to grave doubt. 

 It seems to be certain, however, that durino; the autumn and 

 winter the more violent paroxysms of the volcano occur, and 

 that in summer the action which takes place is far more 

 equable and regular." This means, I take it, that during 

 that portion of the year exposed to the greatest fluctuations 

 of atmospheric pressure, the volcano exhibits the greatest 

 variety in the phases of its eruptions. 



That the Tana volcano is affected in this way, is open to 

 very little doubt at all. 



Almost every visitor from Captain Cook onward has 

 noticed and recorded this fact. I had personally the 

 opportunity of observing the eruptions during all kinds of 

 weather, and the increase in their rapidity and fores was 

 invai-iable when the weather became unsettled, stormy 

 weather approached, on the fall of rain and for a short time 

 after rain. The most violent eruptions have always occurred 

 during the summer, that is the hurricane months ; in the 

 winter when the trade winds are blowing, and day after daj^ 

 the weather does not vary, the eruptions are very regular 

 and comparatively feeble. Mr. Neilson after an experience 

 of its habits for 1 4 years reports in the same terms. 



Although a series of accurate observations of the eruptions 

 taken in connection with those of a barometer would be of 

 great value in throwing light on this subject, the general 

 law of the connection of the condition of the atmosphere and 

 the nature of the eruptions may be considered as established, 

 so far as this volcano is concerned. 



It would be impossible to study the working of this 

 remarkable engine of nature for any length of time without 

 being led to consider the cause of it all, the reason of its regu- 

 larity and of its irregularities, and the various peculiarities 

 attending it. The inquiry is full of interest but beset with 

 difficulties, so that theories must be advanced and received 

 with caution in the limited state of our pi-esent know- 

 ledge. 



1 will, however, conclude this paper with an attempt to 

 account for the varied and remarkable performances of the 

 volcano under discussion. 



The generally accepted theory of volcanic action is that it 

 is due to three causes : — 1st. The existence of molten lava 

 beneath the surface of the earth's crust; 'Jud. The percolation 



G 



