The thermometers that are employed in meteorological observations are generally those made with mercury; for this metal does not freeze except at very low temperatures, such as are never observed in our climates. But, for a wintering in high latitudes, spirit thermometers must be provided. If the freezing and the boiling points are determined on such an instrument, and the interval is divided into 80 or 100 equal parts, its variation will not be parallel to those of a mercurial thermometer; this is due to the unequal dilatation of the two liquids. Therefore, in comparing a spirit with a mercurial thermometer, a correction must be made through the whole extent of the scale.

It is often essential to know the highest and the lowest degrees marked by the thermometer, in a certain determined interval of time. To effect this we have re

course to thermometrographs. Of the different arrangements that have been given, the following is the simplest, and perhaps the best. If a mercurial thermometer is placed horizontally, and a movable index of iron wire or of glass is in contact with the extremity of the mercurial column, this index will be thrust forward as the mercury expands, but will remain at the same place if it contracts under the influence of a reduction of temperature. The point of the scale where the index is found indicates, therefore, the greatest degree of heat, or the maximum to which the instrument has been subjected. An analogous mechanism will give us the lowest degree to which the instrument has fallen. At the extremity of the column of a spirit thermometer a glass index is so placed that it is entirely plunged into the liquid. When the alcohol contracts, it draws with it the index, in consequence of its adhesion to the glass, but, in expanding, it does not displace it; so that the degree, corresponding to the point where the index is found, indicates the minimum, or lowest temperature, to which the instrument has been exposed. If, then, we wish to know the maximum and the minimum of the temperature in the twenty-four hours, the index of iron wire is placed in contact with the mercury, and the upper extremity of the glass index is made to coincide with the extremity of the column of spirit, by inclining the two instruments. The following day the position of the

* Indices of glass, and even those of iron wire, in the course of time, are drowned, that is, they penetrate into the mercury. To avoid this inconvenience, M. GREINER crowns the thermometric column with a cap of very thin glass, by which the mercury is kept from the index.-M.

two indices will give the maximum and the minimum of temperature.*

PROPAGATION OF HEAT. -A body, when in the neighbourhood of another, the temperature of which is higher or lower, becomes heated or cooled. How does this exchange of heat come about? In two ways, as we have learned from experiments-conductibility and radiation.

CONDUCTIBILITY. - If cavities are fashioned in a metal bar, and thermometers are introduced into them, it will be found that, at the end of a certain time, all the thermometers will indicate the same temperature; but, if a lamp is brought near to one end of the bar, this end will be heated first, and, in a very short time, the thermometers will rise, and the more so as they are nearer to the source of heat. The propagation of heat takes place because each layer of metal communicates a portion of its heat to the layer with which it is in contact. This property of bodies to transmit their heat thus, by the intervention of their molecules, is termed conductibility. Conducting power is different in the various bodies in nature. If bars of the same length and thickness, but composed of different materials, are compared, we discover that they do not conduct beat equally well. Metals transmit it very well, then come rocks, afterwards wood, &c. In general, the more porous a substance is, the less conducting power does it possess. A crowd of phenomena, falling under our daily observation, depend simply on the difference of conducting powers. On a summer's day expose to the sun a mass of metal and a mass of wood, of the same volume, and covered with the same varnish; then touch each of them with the hand: the wood, which is hot on the surface only, will at first give the sensation of a hotter body than the metal; the latter, on the contrary, being deeply penetrated by the heat, will produce a less intense sensation at first, but it will continue much longer, because it will gradually transmit to the hand all the heat that it has absorbed. For the same reason, a piece of metal seems much colder in winter than a piece of wood, because the heat of the hand penetrates much more quickly into the metal than into the

* The minimum degree is indicated by the upper extremity, the maximum by the lower extremity of the index.

Six's thermometrograph, modified by BELLANI and BUNTEN, is a more convenient instrument, and more exact than RUTHERFORD'S, which has just been described (vide PECLET, Traité de Physique, t. i. p. 542; and POUIL LET, Elemens de Physique, t. ii. fig. 372). But the indications of this instrument are exact only so long as it has not undergone shocks or pressures, for these displace the index and falsify their results.-M.

wood; the surface alone of the latter becomes heated, and this, too, in a very short space of time. Sand, which is a very bad conductor of heat, is intensely hot on the surface during summer, but, at a few inches deep, this elevated temperature ceases to exist.

RADIATION. Conductibility is not the only means by which bodies interchange temperature. If, in winter, we are at a certain distance from a hot stove, we feel that it transmits heat to us; is this an effect of the conducting power of the air? By no means; for if we place between ourselves and this stove a metal screen, we shall no longer feel the heat, however thin the interposed plate may be; but, as metals are good conductors, the heat ought to traverse the screen. To render this fact perfectly clear, let us place a convex mirror at some distance from the stove, and the bulb of a thermometer in the focus of this mirror. If we place a screen between the mirror and thermometer, the latter will not be affected, but the instant we take away the screen the thermometer rises. Things occur exactly as if the rays of the sun fell on the mirror, and we must admit calorific rays for the same reasons that have induced us to conceive luminous rays. The mode of transmitting heat is termed radiation. Calorific rays easily traverse a certain number of bodies, and especially pure air.

All bodies in nature are incessantly radiating one to another; hence arises a continual interchange of temperature, because some absorb what the others lose by radiation. It is principally the surface of bodies that radiates, and, generally, with greater facility the less polished it is. These losses of heat are partially compensated by the heat transmitted from within outward. If, therefore, we surround bodies that are bad conductors but good radiators, such as swan's down, locks of wool, feathers, light sand, glass, snow, &c., with a very cold atmosphere, and compare their cooling with that of bodies which scarcely radiate, but which are good conductors, such as polished metals, we shall perceive that the former become cold much faster than the latter.

CAPACITY OF BODIES FOR HEAT.-The thermometer teaches us whether bodies, heated either by direct transmission or by radiation, have the same temperature. To render our explanation more comprehensible, suppose the heat to be something material. It is natural to ask if two bodies of the same temperature always possess the same quantity of heat. In other words, is as much heat required to make altemperature of a mass of water rise from 8° to 40° as

to raise the temperature of a mass of iron to the same number of degrees? Experiments answer this question in the negative. A different quantity of heat is required to give an equal change of temperature to bodies of a different nature. In order to have a point of comparison, a kilogramme of water at zero is chosen as unity, and we examine how much heat is required to elevate its temperature 1° C. Then this quantity is determined for other bodies by operating, in like manner, upon a kilogramme of the substance. The quantities found are called the specific heats, and this property obtains the name of the calorific capacity of bodies.

The following experiment shews the truth of what we have said, and explains the method employed to estimate the specific heat of bodies. Pour into a vessel, having thin sides, 500 grammes of water at zero, then add 500 grammes of water at 40°; on making this mixture, we shall have one kilogramme of water at 20°. If the experiment is varied, whatever are the initial temperatures of the two quantities of water, the mixture will always have a temperature equal to half the difference of these initial temperatures.

But if we throw into 500 grammes of water at zero, 500 grammes of iron filings at 40°, the temperature of the mixture will be only 3,96. Thus, then, the 36°,04 of heat, which the iron has lost, have not been able to elevate the temperature of the water more than 3°,96; and iron requires less heat than water to attain the same degree of temperature in the proportion of 3,096, 36°,04. The calorific capacity of water being 1, that of iron will, therefore, be 0,11. The difference is analogous to that existing between bodies in respect to their weight. If we fill flasks of the same capacity with different liquids, such as water, alcohol, mercury, &c., we find great differences of weight. Thus the same volume of mercury will be thirteen times heavier than an equal volume of water. Bodies may be regarded as flasks, into which we have poured heat. The thermometer indicates the same temperature, but just as the weights of equal volumes of these liquids are different, so also bodies, in which the thermometer indicates the same temperature, are possessed of very different quantities of heat. Philosophers have given the name of specific gravity to this inequality of weight, observed in bodies of the same volume; they have also called the unequal capacity of bodies for heat, specific heat.

INFLUENCE OF THE SUN.-The study of the laws that regulate the variations of the temperature of the atmosphere prove that the sun is the principal cause. In pro

portion as this body rises above the horizon, the heat increases; it diminishes as soon as it is set. The differences between summer and winter depend also on the time that it remains below the horizon, and on its distance from the zenith of the observer. Astronomy teaches us, it is true, that the earth was formerly an incandescent globe, which, when launched into space, gradually cooled. In proportion as we descend into the bowels of the earth, we find a greater or less elevation of temperature, which renders the existence of a central fire or incandescent nucleus very probable. But the surface of the earth is composed of bodies that are such bad conductors, that this central heat is very slowly communicated to the atmosphere; and the researches of Fourier have shewn that it may be entirely neglected in Meteorology.

The height of the sun above the horizon is one of the most important elements in the study of its calorific action. In fact, a surface is more highly heated by a distant source of heat, as the line drawn from this source to the surface approaches nearer to the perpendicular. Take an open book, and present it to the light of a lamp on holding it vertically, you will easily be able to read the characters; but, the more you incline it, that is, the smaller the angle is which the incident rays make with the surface of the book, the less will the surface be enlightened, and the more difficult will it be to read the characters; and they will finally become entirely invisible.

Mathematicians have endeavoured to deduce the changes of temperature of days and seasons from the height of the sun; but the action of this body is modified by so many accidental circumstances, that we must have recourse to direct experiment. This is done by means of a thermometer exposed in the open air, on the north side of a building at three or four decimetres from the wall, and at a distance from any white surface likely to reflect the heat. If it becomes wetted by rain, it must be wiped about five minutes before making the observation; for the drops of water, by their evaporation, will lower the temperature of the bulb of the thermometer. Care must be taken, in winter time, that the thermometer is not subjected to a current of hot air coming out from the apartment.

If an instrument thus placed is carefully followed, it will be observed that the temperature changes every moment. It would be an easy matter to turn all these isolated indications to account, in order to compare the temperatures of different months or days in the year. These comparisons