After the operation is finished a little air is sent into Q in such a manner as to sensibly lower the level of the solution of potash, and the upper extremity of R is closed with a rubber stopper. Afterward, the apparatus can be transported without any danger of the potash becoming engaged in the tube K and reaching the measuring tank A. To proceed to the analysis, a stake is driven into the soil to which all of the apparatus can be fixed. At the side of the stake the apparatus for taking the sample, already described, is driven into the soil and this apparatus is connected by the tube N with the apparatus for determining the carbon dioxid. The pinch-cocks L and M being closed, F is lifted until the mercury which runs from it fills A and approaches D. During this time the air which the apparatus contains has been driven out through O. The tube NGD is freed from air by opening the pinch-cock L, lowering F and drawing into A the gas coming from the soil; afterward closing L and driving out the gas through O. After two or three rinsings of this kind, which employ altogether only ten to twelve cubic centimeters, the gas which is to be analyzed is sucked into A. For this purpose F is lowered until the mercury in the tube BC is very near C. The pinch-cock L is closed and M opened. The reservoir F is displaced little by little by pressing lightly against the rectangular board in order to give it greater firmness in such a way as to fix the level of the mercury exactly at C, and the line is noticed where the solution of potash in K stands. The gas contained in the apparatus is under a pressure, the difference of which from the external pressure is represented by the column of the potash solution between the mark just noticed and the level of the same solution in the bulb R. In order to absorb the carbon dioxid, F is lifted until the mercury stands between D and E. The gas thus passes from A into Q. It gives up immediately its carbon dioxid to the potash solution. It is then made to come again into A, and afterward a second time into Q in order to free it from the last trace of dioxid. Finally it is made to return to A and F is kept at such a height that the potash solution maintains in the tube K the same level as at the commencement of the operation. The gas is then at the same pressure to which it was subjected before absorption. The level of the mercury is then read on BC. At the time the apparatus is used, the measuring tube A should be slightly moist. If it is not so, a small quantity of water should be introduced which is afterward rejected, but which leaves a sufficient quantity of moisture upon the internal walls of A. In this way the gas will always, before or after absorption of carbon dioxid, be saturated with vapor of water, and the figure read in the last place upon the tube BC represents the percentage of carbon dioxid in 100 parts of the gas extracted from the soil supposed to be