the thermometer, is obscure, but it is certain that Galileo in Padua used an air-thermometer in the latter part of the sixteenth century, which Rey, a French physician, filled with liquid in 1631. This thermometer, as well as other physical instruments, was perfected by members of the Accademia del Cimento at Florence. These instruments are described in Saggi di Naturali Esperienza, written in 1666, and translated into Latin and English. The Florentine thermometers had one fixed point, that of freezing water, and contained either spirits or mercury. In 1724 Fahrenheit, in Danzig, fixed three points on the scale of the mercurial thermometer, viz. the cold produced by ice and salammoniac which he called o°, freezing water or 32°, and the heat of the human blood which he assumed to be 96°. This thermometric scale, having 180° between freezing and boiling water, and that of Celsius, with 100°, are the only ones in scientific use to-day. It is a remarkable fact in the history of thermometers that neither of these thermometers remained in the country where it was invented; thus the thermometer of Fahrenheit, a German, came into use exclusively in England and her colonies, while that of Celsius, a Swede, is now used on the continent of Europe except in Germany, where the thermometer of Réaumur, a Frenchman, is still in popular use. Of the four fundamental

meteorological instruments, the barometer was the last invented. Aristotle had suspected that air had weight, but it was not demonstrated until the middle of the seventeenth century, when the old axiom "that Nature abhors a vacuum was replaced by the rational explanation, given by Galileo and Torricelli, his pupil, why water will not rise in a suction pump more than thirty-two feet. In 1643 Torricelli executed this famous experiment: he took a glass tube, sealed at one end, and filled it with mercury, then, closing the open end with his finger, he inverted it in a basin of mercury. The mercury fell to about thirty inches, which was recognized to be the weight of a column having the area of the tube and of the height of the atmosphere. The application of the barometer was due to Blaise Pascal, who repeated at Rouen Torricelli's experiment with a much longer tube filled with water, which being thirteen times lighter than mercury, stood thirteen times higher, or thirty-two feet, in the tube. Pascal, being himself at Paris in 1648, got his brother-in-law Perier to carry a barometric tube filled with mercury to the top of the Puy de Dôme, a mountain in Auvergne rising about 3500 feet above the city of Clermont. The mercury fell in the tube with the ascent, and at the top of the mountain it stood some three inches lower than at the base, showing that the

lower layers of the atmosphere are denser than the upper. Pascal repeated the experiment on the Tower of St. Jacques in Paris, and it is interesting to note that more than two hundred years afterwards, meteorological stations were established both there and on the Puy de Dôme. It was soon perceived that not only did the level of the mercury in the tube change with height, but that it oscillated continually at the same place, and from its observed relation to the state of the weather its name "weather-glass" is derived. In 1650 the weight of the air was demonstrated in another manner by Otto von Guericke, burgomaster of Magdeburg, who by means of an air-pump of his invention performed the experiment, which Aristotle had tried unsuccessfully, of weighing a vessel full of air and the same vessel exhausted of air. He also showed the pressure of the air in all directions by the famous experiment of the Magdeburg hemispheres, which, being hollow, were placed together, and after the air was exhausted from the sphere so formed sixteen horses were unable to pull them apart. Soon afterwards Robert Boyle experimented further upon the weight and "spring of the air," as he called it, and gave the name to the barometer. Both Boyle in England and Mariotte in France discovered the law, bearing indifferently their names, that the pressure of gases is propor

C

tional to their density. Halley, a few years later, showed that the rate of decrease in pressure differed from the rate of increase in height, and developed formulæ for measuring heights by the barometer, which were afterwards perfected by Laplace. Knowing the heights of the barometer at a high and at a low-level station, and the mean temperature of the air lying between them, it is possible to compute accurately the difference of height of the two stations, or, conversely, given this height, the difference in barometric pressure can be calculated. By the middle of the seventeenth century the most important meteorological instruments had been invented, and not only can Italy claim to be their birthplace, but the Grand Duke Ferdinand II., whose brother Leopold founded the Accademia del Cimento, distributed the new instruments in Italy and even beyond the Alps, so that in 1654 observations several times a day were begun at a dozen stations. The observations in Florence from 1650 to 1670 were preserved and constitute the commencement of instrumental meteorology.

It was the conquest of Peru which, by leading men over the high passes of the Andes, first brought them to great heights, but although we find mention in the history of the expeditions of the so-called mountain sickness, caused by fatigue

B

as well as by cold and rarefied air, it does not appear that scientific observations were made. Therefore, while it must be assumed that the atmospheric conditions at considerable altitudes were familiar to travellers, yet not until the middle of the last century did Bouguer, one of three French Academicians sent to Peru on a geodetic mission, fix the height of the freezing point in various latitudes, after observing that the temperature fell below freezing at night upon the mountains near the equator. During the latter part of the century, Kirwan, an English chemist, calculated the temperature for various parallels of latitude, and in 1817 Alexander von Humboldt, after a voyage around the world, published his isothermal lines, or lines of equal temperature on the surface of the globe, by which he showed that the deviation from the normal, or calculated, temperature arose from the distribution of land and water, and from the geographical relief of the former. This work of von Humboldt formed the basis of all subsequent studies in comparative climatology. Meanwhile chemistry had kept pace with physics, and in 1774 the old theory, that air was one of the four elements from which all things originated, was rendered untenable by Priestley, who proved that oxygen gas, which he discovered, was a constituent part of air. The other constituent, nitrogen, formerly