\ 
72 REPORT—1850. 
phenomena. Hence, perhaps, but a few instances in which earthquakes 
followed great heat, sufficed to produce amongst the Greeks and Romans 
the opinion that earthquakes occurred more rarely in the cold season of the 
year than in the warm. How little this opinion of the older naturalists is 
founded in reality, may be gathered from the fact that directly the contrary is 
asserted and believed by those of more modern date, namely, that a greater 
number of earthquakes happen in winter than insummer. In different coun- 
tries also different opinions have been entertained on this subject, which per- 
haps always arose from some few, but great, and therefore striking phe- 
nomena. 
5th. The Rain-gauge. 
Torrents of rain have often been noticed as falling during earthquakes, 
and they have also often begun in heavy rain, and sometimes have been 
concluded with rain, and this has in each case often been accompanied 
with thunder, lightning and wind; but, on the other hand, so many earth- 
quakes have occurred with serene skies, before, during, and after the shocks, 
that we must conclude there is no necessary connexion established be- 
tween them. I can find no numerical observations as to rain, in relation 
to our subject, recorded. 
As secondary effects after earthquakes, disturbances in the usual fall of 
rain may be almost certainly anticipated in a degree greater as we approach 
the voleanie centre; but this branch of seismo-meteorology is as yet un- 
touched. 
6th. The Electrometer. 
Eandi observed a Volta-electrometer much agitated during the long-con- 
tinued earthquake movements of Pignorol in 1808 (Journ. de Phys., t. xlvii. 
p- 291.); but even were these experiments more copious and refined, it by no 
means follows that agitations of instruments, at all times in activity, and 
whose extraordinary activity at any moment may depend upon a passing 
cloud, have any necessary connection with earthquakes. 
*« Thunder-storms,” says Von Hoff, “ have undoubtedly on some occasions 
burst forth at the same time with earthquakes. Examples of this are to be 
found in his Chronicle in the years 365, 1138, 1570, 1627, 1680, 1704, 
1711, 1715, 1720, 1752, 1821, 1824, 1828. But how many earthquakes 
took place during this lapse of time without the occurrence of storms, and 
what innumerable storms to the production of which no earthquake contri- 
buted!” (Gesch. Veran. Erdober, Th. iv.) 
On the theory of probabilities only, it would be strange were it not so, 
amidst the numbers of earthquakes of which some record exists. 
But here again we are without any facts (to be truly so called) as respects 
regions visited by earthquakes far beyond the range of the volcanic centres ; 
within these, or in proportion as they are approached, of course this, like 
other atmospheric perturbations, may be expected. 
7th. The Magnetometer. 
Humboldt found, in the great earthquake of Cumana (4th Nov. 1799), 
the declination and magnetic intensity unaffected, but to his surprise the dip 
was diminished by 48 minutes. He had no ground to suppose an error. 
With this solitary exception, in all the other earthquakes he experienced 
on the high lands of Quito and Lima, all the magnetic elements remained 
unaffected. (Relat. Hist., t. i. pp. 515, 517.) 
These movements are of course totally different to those which have been 
observed by Arago, Biot, and repeatedly at Dublin by Dr. Lloyd, in their 
magnetic observatories, viz. oscillations suddenly affecting the magnetometers, 
and most probably due to the transmission of very small impulses from 
