T. H. Streets—Earthquakes in Japan. 367 
ll rainy days; in August, of the same year, when there were 
two shocks, there was a rainfall of but ‘92 inch. In Septem- 
t 
on the 3d of the same month, and a rainfall of 11°99 inches. 
It rained continuously from the 20th to the 30th, but no more 
shocks occurred until October 25th. March is likewise a wet 
month, but the amount of rainfall is less than in September. 
In voleanic regions the rain is thought to supply steam, which 
equinox and the time of the greatest frequency of earthquake 
shocks, the one occurring in August and Sep ( 
Other in February and March, being separated by a period of 
four months either way. I am aware that this conclusion 
differs from the opinion usually held, which is, that earth- 
quakes are associated with the equinoctial period of atmos- 
‘There occurred two shocks sufficiently severe to do damage 
December 
that the country has ex rienced since 1850. 
period is observed, oer i uch a period has been stated by 
st. 
Some to exi 
