TRANSACTIONS OF THE SECTIONS. 83 
that the earth had opened and destroyed various animals, and that Talavera with 
other places had suffered. 
On the 28th of June, 1847,a severe earthquake happened at.Ica, which continued 
at intervals during two days. The shocks were both vibratory and vertical, and did 
much damage to the town: one writer complained of heavy loss from the fracture of 
large jars in the ground which contained liquors. Tacna, 10th September, 1847, 
10 p.m.: rain has fallen all day, which is rare here; also the sun has not been seen 
since yesterday; barometer has fallen th since morning, night very dark. 11th, 
3 a.m. : awakened by a violent earthquake, which lasted 1 think half'a minute; others 
say it continued two or three minutes. The motion was both vertical and in oscil- 
lations, with land noises more like a succession of reports or explosions, than the 
usual subterranean rumble; little damage was done beyond cracking some walls; 
immediately I examined the barometer, which was as at 10 p.m., rain still falling, 
atmosphere thick and heavy, sand in tube fallen. 
It appears from Ee Comercio of Lima, of 8th October, that two strong shocks of 
earthquake were felt at Arequipa, on the I 1th ultimo, one at five minutes before 
3 a.M., ὦ. 6. about the time of that above noted at Tacna and Arica; the distance ina 
direct line from Arequipa to Arica is about 200 miles. 
On the 8th of October, 1847, an earthquake was felt throughout Chili, from N. to S., 
and more severely at Melepilla, which is between Valparaiso and Santiago ; at Mele- 
pilla, the earth was shaking or swinging during two days, within which time several 
hundred shocks were experienced, the convulsion having been the worst here since 
that of 1822. 
Though no certain indications of an approaching earthquake can be noted, yet I 
have heard persons in Arica and Tacna affirm that previous to heavy shocks, they 
were often sensible of a peculiar disagreeable smell or state of the atmosphere, which 
they assert may be considered a precursor of a coming shock. Between 1843-48, the 
said odour was more rarely referred to in that quarter, which may have resulted from 
the greater paucity of earthquakes there within that period. In 1826 and following 
years, prior to some severe shocks there, my olfactory organs were affected by the 
invisible agent above noted, which was something different from the effluvia emana- 
ting from decaying animal and vegetable matters. Native Peruvians call it ‘Olor de 
tierra,’ smell of the earth. 
Another indication of a coming earthquake, in the opinion of the people there, is, 
that when the atmosphere is profoundly calm or stagnant, an almost imperceptible 
movement of it, a breath, as it were, of wind is felt, so gentle as to affect only the 
faces of persons who are peculiarly sensitive, These, and other alleged precursors 
of severe earthquakes, such as movements of the lower animals, should he diligently 
watched by those who would study these so often dire visitations ; which, though no 
human power can avert, yet perchance may, in some cases, be made less destructive ; 
as, with the aid of properly adapted instruments and other means, attentive observers 
may obtain, not only a solution of the problem as to the causes of these terrestrial 
convulsions, but also a knowledge of their premonitory symptoms, if any, which 
would be of vast importance as affecting precautionary measures for saving life and 
property. 
On the Position of the Footsteps in the Bunter Sandstone of Dumfries-shire. 
By Rosert Harkness. 
The Bunter sandstone occupies four separate portions of the county of Dumfries : 
one occurring in the north-west, in which as yet no footprints have been discovered ; 
another is found filling up the flat parts of the Vale of the Annan; the third is met 
around Dumfries ; and the fourth stretches east from near the village of Cummer- 
trees along the shore of the Solway, and occupies the lower part of the parish of 
Cannobie. It is in that portion which lies in the low part of the Vale of the Annan 
that impressions of footsteps occur in the greatest abundance. At Corncockle, and 
also at Templand quarries in this district, and likewise at Locherbriggs, Craigs, and 
Green Mill, in the sandstone of Dumfries, footsteps are met with. At the former of 
these places they are found more commonly and in greater variety, as well as in 
better state of preservation than at the other localities. The nature of the sandstone 
which affords the tracts, is to a great extent similar; and the direction of the dip is 
G2 
