MEAN TEMPERATURES OF THE DIFFERENT HOURS AT LEITH, NEAR EDINBURGH. HOURS. JAN. FEB. MARCH. APRIL. MAY. JUNE. JULY. AUG. SEP. ост. NOV. DEC. HOURS. 7,62 4,90 4,25 4,41 7,23 4,83 4,19 4,12 ,62 Midnight 4,79 4,18 4,04 6,34 13 ,79 4,26 3,86 6,16 6o,29 9,86 5,44 | 4,41 9,14 4,90 10°,01 11°,43 14°,81| 17°,69 160,26 15°,28 10,85 8,80 4,263,87 4,06 9 16 3,84 3,89 19 9,00 4,50 3,96 21 4,66 4,31 4,39 7,98 9,85 13,36 15,83 14,88 13,62 9,37 4,76 4,09 21 Means. 5°,00 4°,74 40,84 4,27 70,83 90,91 130°,26 15°,70 14°,60 | 13°,54 9°,54 5°,07 4°,26 Means. 2223 20 These tables shew that there is a maximum and a minimum of temperature on each day. The minimum occurs some little time before the rising of the sun, the maximum about two o'clock in the afternoon; a little sooner in winter and a little later in summer. The majority of philosophers admit that the moment of the rising of the sun is that at which the temperature is the lowest; but if we deduce from observations a formula, independent of the trifling errors of reading, which are almost inevitable, we shall find that the minimum occurs about half an hour before the rising of the sun, at the time when this body is yet 12° beneath the horizon. This rule, which is only applicable to our climates, varies in different seasons. In autumn and in winter the minimum coincides with a depression of 18° below the horizon, and in summer with 6° only. When the sun is above the horizon, it acts upon the earth and the lower strata of the atmosphere with greater power, as its angular height is greater. One portion of this heat penetrates into the soil, the other is lost by radiation towards the atmosphere and celestial space. Before mid-day the earth receives, in every instant of time, a quantity of heat, exceeding that which it loses by radiation; and its temperature is raised. This effect continues also for some time after the sun has passed the meridian, hence it follows that the maximum takes place some hours after the time of noon.* While the sun is sinking toward the horizon, its We know that, during the winter at the polar regions, the sun entirely ceases to appear above the horizon for a period of time, which is longer as the observer is nearer the pole. The variations of the thermometer are then due, almost entirely, to the state of the sky and the direction of the wind. We may, therefore, expect to find an almost insensible daily variation. The following are the results which were obtained, during the winter of 1838-9, at Bosekop, in Lapland, under the 70th degree of latitude, by MM. LOTTIN, BRAVAIS, LILLIEHOOK, and SILJESTROEM, members of the Commission of the North. They confined themselves to the mean temperature of the forty days preceding and the forty days succeeding the winter solstice : If entire confidence is to be placed in mere numbers, deduced from the action becomes less powerful, and the loss by radiation exceeds the gain by absorption. The heat diminishes the more rapidly as the sun is nearer setting. As soon as it has disappeared, the calorific source no longer existing, all the acquired heat radiates towards celestial space, the temperature falls, and would fall still lower, if the portion of the heat, which had penetrated into the superficial layers of the soil, did not return to the surface, by virtue of the conducting power of the earth. The lowering of temperature continues until morning announces the return of the sun, which again heats the regions that it illuminates. These explanations are so simple, that almost all the phenomena of temperature may be deduced from them; and we may also infer from them that the diurnal variation of heat is, probably, the same in all countries. However, in hot climates on the sea-coast, the maximum of temperature often occurs before the arrival of the sun at the meridian; because there arises, about mid-day, a fresh breeze from the sea, which lowers the temperature. Daily observations, made in July and August, 1837, on the borders of the Baltic, have shewn me that the maximum takes place in the afternoon; although several hours sooner than at Halle, which is in the interior of the Continent.* observations made during one winter only, we should be led to admit that the temperature attains its maximum about eleven o'clock, A.M., its minimum about five o'clock, P.M., and a second maximum at eleven o'clock, P.M. But these two oscillations are not sufficiently marked to authorise us as yet in vouching for their constancy. It would, however, be interesting to establish the existence of a daily calorific wave, which should not depend on the direct action of the solar rays. The preceding observations shew that its amplitude is not greater than 0°,4. It remains to be proved whether this phenomenon may be explained by a regular diurnal change in the state of the heavens or in the direction of the wind, by the diurnal variation in the pressure of the air, by aurora borealis, or by any other local or general influence.-M. * On the summit of high mountains different causes produce an analogous effect. The hottest period of the day occurs half an hour, or threequarters of an hour, after the term of the culmination of the sun. M. KAEMTZ'S observations on the Faulhorn (canton of Berne) at 2683 metres above the level of the sea, in 1832 and 1833; those which I made, with MM. BRAVAIS and WACHSMUTH, on the same mountain in 1841; and, finally, those of MM. PELTIER and BRAVAIS in 1842, all agree in giving the same result, however various these different series may have been in respect to the meteorological circumstances which accompanied them. The observations on the Great St. Bernard also prove that, in this station, the temperature of mid-day is generally greater than that of three o'clock. If, for these two stations, the period of maximum differs, this difference is very probably due to Faulhorn's being an isolated summit, whilst the hospice of St. Bernard is overlooked on all sides by elevated summits. The coldest period coincides with that of the plain, that is, it happens about half an hour before the rising of the sun. With respect to the diurnal variation of the thermometer, the climate of the summit of mountains bears a great analogy to sea-climates.-M. 1 DETERMINATION OF MEAN TEMPERATURE.— Twice during the day, the degree of the thermometer is equal to the mean temperature. It might, therefore, seem that it would be sufficient for us to take an observation at one of these two periods. But this method is very unworthy of trust. At the moment when it oscillates about the mean, the temperature changes very rapidly; and, if the observation is made a little too soon or a little too late, very notable errors will result. The surer way is to read it off several times during the day, and at such hours that their arithmetical mean shall approach the true mean as nearly as possible. By taking observations at 4 o'clock and at 10 A.M., and at 4 and 10 o'clock P.M., the fourth of the sum of the temperatures found will give a value that will differ very little from that of the mean. The arithmetical mean of observations at 6 A.M.,2 P.M.. and 10 P.M., also differs very little. M. Schouw, in order to facilitate the labour of observers, recommends their confining themselves to three readings,-at 7 A.M., at noon, and at 10 P.M.; the sum of the degrees divided by 3 will give the mean.* At Halle, the mean of June, obtained by this method, is 150,95; whilst the true mean is 15,72; consequently 0°,23 lower. If the true mean at Halle were not known, we might arrive at it by determining, by means of observations made in another country, the quantity that must be deducted from the mean deduced from three observations only. Thus, at Padua, the mean found by this latter method is 22°,23; the true mean is 21°,93, the difference=0°,30. Deduct this difference from the mean deduced at Halle from three observations only, and we have 15°,95 — 0°,30—15o,65; a number which is not very far from the true mean, 15°,72. In fine, we must be sure as to whether the arithmetical mean of the three observations is greater or less than the true mean, in the places where the latter is known, and add or deduct this difference from the mean obtained in the place where the observations are being made. For points situate on the Continent, and in the temperate zones, we should take the means of Göttingen, Halle, and Padua, in order to eliminate the little anomalies which might exist in one of these cities taken separately. For England, we should adopt the correction found by the observations made at Leith. Although we do not attain to results that are perfectly accurate, they * In general, the mean of four readings, made at equidistant hours, is not far from the true mean. will yet be much nearer the truth than those that are not corrected. The mean being intermediate between the maximum and minimum of the day, it has often been proposed to deduce it from these two elements, and to consider the half of their sum as the true mean. But the latter is widely different from this empirical mean; and here again we should be obliged to have recourse to a correction which experiment alone can disclose to us. This correction is a constant coefficient, by which the excess of the maximum over the minimum is to be multiplied; the product is added to the minimum; the sum is the true mean sought after. This coefficient varies according to the process by which the maximum and minimum have been obtained. If a common thermometer is observed at the hours of maximum and minimum, the readings will differ from the indications of the thermometrograph. In fact, this instrument always indicates the true maximum and minimum; whilst the former method does not give them, because the atmospheric accidents of the day may displace the maximum and minimum by several hours. The mean found by the thermometrograph will, therefore, differ from that deduced from direct observation at the presumed periods of diurnal maximum and minimum ; because the maximum will be always higher, and the minimum lower, than those of the thermometrograph. In many cities, such as Paris, Brussels, and Basle, observations of the thermometrograph and the thermometers have been simultaneously taken several times during the day. We may, therefore, declare, from these comparative observations, a coefficient by means of which we may determine the true mean. With the series of Padua, Halle, Göttingen, and Leith, we have calculated the factor that will enable us to deduce the true mean temperature from the periods of the maximum and minimum. The following table gives these two coefficients for all the months of the year : |