50 REPORT—1858. 
Within the first period (in the 18th century) we find the great Lisbon 
earthquake; within the second, in the same century, the great Calabrian 
one. We find (referring to the Catalogue itself) earthquakes in great num- 
bers, and many great ones—in the Mediterranean basin in the middle of 
the 17th century, and the great Jamaica earthquake in its latter decade; 
and in the 16th century, its middle period was marked by great earth- 
quakes in China and in Europe, and the latter period by numerous shocks, 
and most of them severe, as at the Azores, &ec. Whether the latter half of 
our century shall show the like, remains to be seen ; from its commencement, 
however, it presents no paroxysmal period comparable to that between 
1840 and 1850. 
While this general resemblance of the curves of these latter centuries 
admits of no doubt, I would forbear from founding anything thereupon be- 
yond this ;—that within this time there seems to elapse a period of about a 
century between each of the very greatest paroxysms (number and intensity 
together) of earthquakes, and a like period between two other consecutive 
paroxysms, of which the second is the next greatest observable, although 
far below the first in power; that a period of thirty to forty years seems 
to occur between the first and very greatest paroxysm, and that next in 
power below it; and that in the middle period (especially in the 17th and 
18th centuries) the number of earthquakes is greatest that crowd into a 
very brief time (four or five years), while at the latter period the number 
is thickly spread over ten or twelve years. 
Upon the whole, the forms of the curves appear to indicate a compara- 
tively sudden burst of seismic energy at each great paroxysm, and (by 
their flat tops or more sloping lines to the right hand) a more gradual 
subsidence, as if the train of causes required time to regain, after one spent 
paroxysm, their energy and regimen, which, when restored, were suddenly 
put into action, and which, once developed, were slow in being wholly 
expended and relapsing into repose. } 
The occurrence of such epochs at the middle, or towards the end of our — 
purely arbitrary subdivision of duration into centuries, must be of course 
only accident. The interval of duration between one epoch and the next, 
is that alone which can have a cosmical basis. 
We may then provisionally affirm the probability of two periods of earth- 
quake maxima—a greater and a less alternately—as occurring in a hundred 
years, for the last three centuries of history at least. The existence of 
some periodic maxima in remoter centuries can hardly be doubted, although 
the epochs of the two maxima have a secular movement, and do not fall in 
the same place in the older times. Anterior to the 16th century, however, 
the general curves of time (Plates I. II. and III.) are, through paucity of 
observations, not sufficiently “ prononcées” to enable this to be asserted from 
them, or to warrant the graphic representation of the epochs of occurrence 
of such paroxysmal periodic maxima for the whole even of the Christian era. 
In Plate VII. fig. 2, the periods of paroxysm (number and intensity) are 
summed and grouped for each successive century of our era. The Ist, 
5th, 9th, 12th, and 18th centuries are those of greatest seismic develop- 
ment, while the Ist and 2nd centuries a.c., and the 3rd, 7th, 10th, and 
14th centuries of our era, are times of comparative repose. The numerical — 
value of the paroxysmal centuries (as we may term them) increases, though 
not regularly, as the present time is neared, and is modified, without doubt, — 
by the same conditions of observation that affect the expansions of the later — 
curves of time. We dare not base any generalization upon it. 4 
Numerically, we find the following average ratios of earthquakes for the _ 
