It is also a matter for surprise that in Paris, among so many sources of carbonic acid, the furnace fires, the respiration of men and animals, and the spontaneous decomposition and decay of organic substances, the quantity of carbonic acid does not exceed 3.5. If, then, the great general mean of normal atmospheric carbonic acid deviates but little from 2.9 or 3.0, it is not doubtful that under local conditions, in closed places, and under exceptional meteorological conditions, considerable variations may occur in these proportions. But these variations do not affect the general laws of the composition of the atmosphere. There are two entirely distinct points from which the measurement of the atmospheric carbonic acid may be contemplated. The first consists in considering it as a geological element which belongs to the gaseous envelope of the earth in general, and it leads us to express the general relation of carbonic acid to the quantity of air, as about three volumes in 10,000. The second, which relates to accidental and local phenomena, to the activity of man and beast, to the effect of fires and of decomposing organic matter, to volcanic emanations, and finally to the action of clouds and rain, permits us to recognize the changes which can occur in air exposed to the influences mentioned, and to a certain extent confined. Without denying that it is of interest from a meteorological and hygienic standpoint, it does not take the same rank as first. J. Reiset's experiments, by their number, accuracy, the large volumes employed, and the interval of years that separate them, have definitely established two facts on which the earth's history must depend: the first is, that the percentage of carbonic acid in the air scarcely changes; the second, that it differs but little from three ten-thousandths by volume. These results are fully confirmed by the results which were obtained by Franz Schulze, in Rostock, in 1868, 1869, 1870, and 1871. The averages which he got, with very small variation, were 2.8668 for 1869, 2.9052 for 1870, and 3.0126 for 1871. More recently Muentz and Aubin have analyzed air collected on the plains near Paris, on the Pic du Midi, and on the top of Puy-de-Dome. Their results agree with those published by Reiset and Schulze. The grand average of carbonic oxide in the air seems to be tolerably fixed, but after this starting-point is established it remains to study the variations that it is capable of, not from local causes, which are of little importance, but from general causes connected with large movements of the air. Upon this study, which demands the co-operation of a definite number of observers stationed at different and distant points of the earth, the experiments being made simultaneously and by comparable methods.