The latter giving a mean of 158.17 m.mrs., against 69.20 m.mrs. in 1857, or a rainfall, for the month and month before the shock, of 88.97 m.mrs. below the average (less than one-half) of the four years anterior; and this is not compensated by the rainfall of the first two months of 1858, although that for February is above the preceding four years' means, and so shows a tendency towards compensation.
The mean temperatures of November and December, 1857, are very near those of the preceding years.
At first sight, this large increment of mean pressure, and deficiency of rain, might induce the supposition that the perturbations are, in some way, directly connected, in the relation of cause and eflfect, with the earthquake; and a more careful discussion of the table leads to the conclusion that such a relation really may exist. On the 12th October, 1856, a shock of earthquake was experienced at the Marine Observatory, which is recorded by the director in the table for that month and year. It was sufficient in power to stop one of the clocks, and the direction of wave-transit is stated to have been in a path from W.N.W. to E. N. E., but was most probably in the contrary direction. The precise meteorological conditions were observed immediately after the shock, which occurred at 1h 58m A.M. Naples mean time, and are given; but present nothing remarkable as to the epoch of occurrence.
If we examine the annual means of rain, however, we shall find that these earthquake years, 1856, 1857, present a decisive deficiency.
The mean of the rainfall for the years 1853, 1854, 1855, is = 1070.4 m.mr. The rainfall for 1856 is but 880.9 m.mr., and for 1857 but 982.6 m.mr., or nearly one-fifth below the
