In Fig. 4 is shewn a fac simile of part of a message received and recorded by a Siphon Recorder, such as is shewn in Fig. 1, from one of the Eastern Telegraph Co.'s Cables of about 830 miles length. W. T.]

ART. LXXXVI. ON THE EFFECTS OF MECHANICAL STRAIN ON THE THERMO-ELECTRIC QUALITIES OF METALS.

HAVING found by experiment that iron and copper wires, when stretched by forces insufficient to cause any permanent elongation, had their thermo-electric qualities altered, but immediately fell back to their primitive condition in this respect when the stretching forces were removed; having remarked that these temporary effects were in each case the reverse of the permanent thermoelectric effects previously discovered by Magnus, as resulting from permanent elongation of the wires, by drawing them through holes in a draw-plate; and thinking it most probable that all these effects depended on mechanical induction of the thermo-electric qualities of a crystal in the metals operated upon; the author undertook an experimental investigation of the thermo-electric effects of mechanical strains, in which he intended to include longitudinal extension, longitudinal compression, lateral compression, and lateral extension, and in each case to test both the temporary effects of strains within the elastic limits of the substance, and the residual alterations in thermo-electric quality manifested after the cessation of the constraining force, when this has been so great as to give the substance a permanent set*. The cycle of experiments has now been so nearly completed for both the temporary and the permanent strains, as to allow the author to conclude with certainty that the peculiar thermo-electric qualities induced in each case are those of a crystal. Thus, he finds that iron bars, hardened by longitudinal compression, have the reverse thermo-electric property to that discovered by Magnus in iron wires hardened by drawing; and that iron wire, under lateral compression, manifests the same thermo-electric property as the author had discovered in iron wire while under a longitudinal stretching force. The apparatus by which these results were

[* See Art. XCI., Part III., below.]

obtained was exhibited to the Section, and the mode of experimenting fully described. As regards iron, the general conclusion is, that its thermo-electric quality, when under pressure in one direction, deviates from that of the unstrained metal, towards bismuth for currents in the direction of the strain, and towards antimony for currents perpendicular to this direction; while for all cases that have been examined, the residual thermo-electric effect of a permanent strain is the reverse of the temporary thermoelectric effect which subsists as long as the constraining force is kept applied. Those of the other metals which have been as yet examined, namely, Copper, Lead, Cadmium, Tin, Zinc, Brass, Steel, and Platinum (specimens supplied as chemically pure by Messrs. Matthey and Johnson being in general used), showed uniformly the reverse effect to that of iron when similarly treated. The effects of permanent lateral compression by hammering were those which were chiefly tested for in this list of metals, and were in almost every case of a very marked and unmistakeable kind. Curious results were also obtained by carefully annealing portions of wires which had been suddenly cooled, and leaving the remaining parts unannealed. Tin and Cadmium thus treated have, as yet, given only doubtful results; Platinum has not been tried; Iron, Steel, Copper, and Brass have given decided indications, in which the unannealed portions showed the same kind of thermo-electric effect as had been found to be produced by permanent lateral compression.

[From Brit. Assoc. Rep., 1855. Part 2.]

ART. LXXXVII. ON THE USE OF OBSERVATIONS OF TERRESTRIAL TEMPERATURE FOR THE INVESTIGATION OF ABSOLUTE DATES IN GEOLOGY.

THE relative thermal conductivities of different substances have been investigated by many experimenters; but the only absolute determinations yet made in this most important subject are due to Professor James Forbes *, who has deduced the absolute thermal conductivity of the trap rock of Calton Hill, of the sandstone of Craigleith Quarry, and of the sand below the soil of the Experimental Gardens, from observations on terrestrial temperature, which were carried on for five years in these three localities (all in the immediate neighbourhood of Edinburgh), by means of thermometers constructed and laid, under his care, by the British Association. The author of the present communication explained briefly a method of reduction depending on elementary formulæ of the theory of the conduction of heat given by the great French mathematician Fourier, which proved to be more complete and satisfactory than the method indicated by Poisson†, which had been adopted by Professor Forbes. He applied it both to the series of observations used by Professor Forbes, and to a continuation of the observations on the trap rock of Calton Hill, which has been carried on up to the present time at the Royal Observatory of Edinburgh, and of which eleven years complete have been supplied to the author in manuscript, through the kindness of Professor Piazzi Smyth. The results, as regards thermal conductivities, show that the determinations originally given by Professor Forbes do not require very considerable corrections; and are satisfactory, inasmuch as values derived from the diminution of the extent of variation of the temperature for the deep thermometers agree very closely with those derived from the retardation of the periods of * "Account of some Experiments on the Temperature of the Earth near Edinburgh," Trans. Roy. Soc. Edinb., Vol. xvI. Part 2.

Poisson's "Théorie Mathématique de la Chaleur," Chap. XII.

summer heat and winter cold at the different depths. They show very decidedly a somewhat greater conductivity of the trap rock at the greater depths (from twelve to twenty-four feet) than between the three feet deep and the six feet, or between the six feet and the twelve feet thermometers, but do not establish any such variation in the properties of the sandstone, and of the sand of the two other localities. A comparison of the mean temperatures of the four thermometers, for the whole sixteen years' observation, shows an increase of indicated temperature in going downwards in Calton Hill, which apparently is much more rapid between the upper than between the lower thermometers; so much so, as not to be referable to the greater conductivity of the rock in the lower position. The author remarked, that, to make the observations available for giving with accuracy the mean absolute temperatures at the different depths, it would be necessary to have the thermometers taken up and re-compared with a standard thermometer. It is most probable that the zero-points of all the thermometers have risen considerably since they were first laid, because the apparent mean temperatures, as shown by the thermometers, are much higher of late than they were at first. Thus, for the period of five years examined by Professor Forbes, and for the succeeding period of eleven years, the means at the different depths are as follows

Notwithstanding the cause of uncertainty which has been alluded to, these results make it highly probable that the augmentation of mean temperature from three feet to twenty-four feet below the surface, apparently 1°38 Fahr. in the first period and •84o in the second period, must be really more than half a degree, or more than the greatest elevation of temperature that had been observed, for a depth of twenty-one feet, in any other part of the earth. The author was struck with this, and reflecting that probably the Edinburgh observations are the only ones that have been made on the interior temperature of other igneous rocks than granite, supposed it to indicate the comparatively modern