Bie (Robern Maule 247 
of earthquakes recorded is stated by Mallet to be 6831, of which 
216 were “ great’ earthquakes, or strong enough to reduce whole 
towns to ruins. Judging from the last 150 years, during which the 
record of such disasters may be supposed complete, Mallet estimates 
that a great earthquake occurs on an average once every eight 
months. The rapid expansion in the number of entries, especially 
during the last three centuries, is no proof, in Mallet’s opinion, of 
any actual increase in earthquake-frequency. He regards it as, in 
fact, “a record of the advance of human enterprise, travel, and 
observation,” the evidence tending rather to the conclusion that 
“during all historic time the amount of seismic energy over the 
observed portions of our globe must have been nearly constant.” 
At the same time, there is clear evidence of irregular and paroxysmal 
outbursts of energy in reference to shorter periods. The frequency 
curves for the last three centuries and a half (1500-1850) show, 
indeed, that while the least interval of repose may be a year or two, 
the average interval is from five to ten years; that the shorter 
intervals are usually connected with periods of diminished earth- 
quake frequency ; that the alternations of paroxysm and repose 
appear to follow no discernible law; except that two marked 
periods of extreme paroxysm occur in each century, the greater 
about the middle and the less not far from the end. 
Still more important are Mallet’s conclusions on the distribution 
of earthquakes in space. Previous seismic maps of the world, such 
as those in the physical atlases of Berghaus and Johnston, were 
based on imperfect catalogues. In neither, as he notices, is there any 
attempt made to depict by various tints the greater or less frequency 
and violence of earthquakes in different areas. Mallet’s materials 
were much fuller, though still far from complete. The original 
map, on Mercator’s projection, is 75 inches long and 48 inches wide 
(reduced in the report to about one-third linear dimensions). He 
divided earthquakes into three classes—great, mean, and minor. 
If their disturbed areas were unknown, he assumed that their radii 
were 540, 180, and 60 geographical miles, and their areas were 
coloured by three tints, the intensities of which were as the numbers 
9,3, and 1. Thus, the most deeply shaded areas on the map repre- 
sent those in which earthquakes are most frequent and violent. 
While its defects are obvious, Mallet’s map remained for nearly 
half a century our best representation of the distribution of earth- 
quakes over the globe. The more important results which he deduced 
from it are the following: as the distribution of earthquakes is 
paroxysmal in time, so also it is local in space; the normal type 
of distribution is in bands of variable and great width (from five to 
fifteen degrees); these bands very generally follow the lines of 
elevation which divide the great oceanic or terr-oceanic basins of 
the earth’s surface; in so far as these are frequently the lines of 
mountain-chains, and these latter those of volcanic vents, so the 
Seismic bands are found to follow them likewise; the regions of 
