92 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
this locality were lost in a shipwreck which I suffered a few weeks 
after these studies were made, I am not in a position to give very 
exact statements as to the numbers of earthquake fountains which 
were formed, but as my notes were as far as possible reproduced 
from memory within two months from the time they were made and 
as this memory agreed with the recollections of a gentleman who 
was with me, they may be accepted as trustworthy. According to 
this record there were twenty-eight pits on the low ground, and on an 
equally extensive area at the side and end of it there were but 
twelve. Those on the higher ground were on an average somewhat 
larger than those on the lower. They were all clearly due to the 
escape of water and not to be confounded with the openings of 
animal burrows and other depressions. The statements of the 
natives were to the effect that they had been formed at the time of 
the shock. Nowhere else in the field of my observations have I ever 
seen these fountain pits anything like as numerous or as well placed 
for study. 
The difference in the development of the earthquake fountains, 
due to the slight difference of level of ground as above noted, clearly 
shows that the seat of the origin of the gas discharges which 
impelled the water upwards was at no great depth below the surface. 
It appears likely that, if the covering had been at any point artifi¬ 
cially made twenty feet or so thicker than it was, the expulsive 
action would not have been great enough in any instance to bring 
the ascending column to the surface. The gas would have been 
lost in the interstices of the material. It is indeed likely that, 
where the porous covering above the level where the gas is 
liberated is thick, the ascending columns may often become 
dissipated before they attain the open air. This in places may 
account for the absence of these pits, although movements which 
might otherwise have produced them actually took place in the 
under rock, the upgoing currents may have been absorbed in the 
porous layer. 
When we seek to examine into the physical history of these shock 
fountains, we find that there is little in the way of observation or 
experiment to guide in the exploration. Some very easily observed 
facts may however serve as a guide: if we watch any fluid which is 
disengaging gas, as for instance champagne in a glass, we note that 
the bubbles gather into vertical streams. The fact is, that, as soon as 
a bubble forms, the pressure of the column immediately over it 
