266 The National Geographic Magazine 



tivity was first observed in the middle 

 of April, 1902. These phenomena were 

 at first slight tremors which shook the 

 dishes on the shelves in the house of 

 Mrs Prentiss, wife of the American 

 consul at St Pierre, and which were un- 

 doubtedly produced by the first audible 

 explosions within the mountain from 

 the ascending column of magma coming 

 in contact with moisture. 



RELATIONS TO THE ERUPTION IN 



ST VINCENT 



The synchronism of this eruption 

 with that of St Vincent, a hundred 

 miles distant, and volcanoes of a similar 

 explosive andesitic character in Central 

 America, to say nothing of disturbances 

 reported in volcanic area throughout the 

 world, is strangely, almost positively, 

 suggestive that the cause of the eruption 

 of Pelee was not the development of a 

 local fissure suddenly letting the water 

 of the sea down to the depths of the hot 

 magma, but, upon the contrary, resulted 

 from a widely occurring disturbance 

 within the interior of the earth's magma, 

 which caused it to rise to meet the upper 

 wet zone, rather than the water of the 

 latter to descend to it, and which is as 

 yet inexplicable. 



IMPORTANCE OF ELECTRIC AND 

 MAGNETIC PHONOMENA 



Finally, accompanying this eruption 

 were phenomena, electricity, magnet- 

 ism, and gases, which, while not prob- 

 ably occurring for the first time, have 

 been conspicuously brought to our atten- 

 tion, and may lead to important deduc- 

 tions upon the origin of magnetic storms 

 and the nature of the earth's interior. 



The fatal cloud came from the lower 

 vent, two miles north of the city. 



The vapors coming from the volcano 

 were and are sulphurous. 



There was force accompanying the 

 eruptions of the morning of Ma} r 8 of 

 great destructive nature, which left 

 much evidence that the gases within 

 the cloud exploded after having reached 

 the air. 



The summit eruptions were and are 

 accompanied by tremendous electric 

 (lightning-like) phenomena. 



A great magnetic storm accompanied 

 the eruptions of Ma}' 8, which was re- 

 corded at remote points, and which indi- 

 cated a connection between them and 

 -the volcanic eruptions. 



In connection with the magnetic phe- 

 nomena, it is of interest to note that 

 Poej' has presented a table showing the 

 relation of Antillean earthquakes and 

 eruptions to the period of sun spots. 

 Of 38 seismic tempests in the Antilles, 

 17 occurred near the maximum sun- 

 spot epochs, and 14 near the minimum. 

 Those of 1846, 1851, 1.852, and 1853 

 were found midway between. The 

 maximum number of volcanic eruptions 

 occur at the minimum of sun spots, and 

 vice versa. 



There is some evidence that flame is 

 emitted with the summit explosions. 



All the evidence indicates that the 

 fatal explosions were not from the site 

 of the old crater, which is five miles dis- 

 tant from St Pierre. All the circum- 

 stantial evidence and much of the direct 

 strongly indicate that the destruction 

 was caused by the eruption from the 

 lower vent, about two miles above the 

 mouth of the Riviere Blanche and some 

 distance north of the city. 



RESUME 



The 30,000 people of St Pierre were 

 exterminated within a few minutes and 

 the town set on fire by the sudden erup- 

 tion of a volcanic cloud. 



* CHOICE OF THEORIES OF THE 

 CATASTROPHE 



The foregoing data and essential facts 

 concerning the great catastrophe I have 

 not attempted to interpret, nor do I wish 



