Nevertheless, this estimation is attended with great difficulty. Not everyone is able to take up such questions, and not all processes are adapted to it. The first thought which would naturally arise would be to inclose a known volume of air in a given vessel, and then determine its carbonic acid by measuring or weighing it. In this way we should obtain the exact relation between a volume of air and the volume of carbonic acid in it, for any given moment, and in any given place. If, however, this be done with a ten-liter flask, for example, it would only hold 3 c.c. of carbonic acid, weighing 6 milligrammes; and, whether it is weighed or measured, the error may easily equal 10 per cent. of the real value, hence no deductions could be drawn from the observed facts. For this reason larger volumes of air were taken, and a current of air, whose volume could be accurately measured by known methods, was passed through condensers capable of retaining the carbonic acid. But in this case the air must pass very slowly through it, so that the process may last several hours; and since the air is continually in motion, owing to vertical and horizontal currents, the experiment may be begun with the air of one place, and concluded with air from a far distant spot. For example, if an experiment lasting twenty-four hours was made in Paris when the air moved but four meters per second (nine or ten miles per hour), it might be begun with air from the Department of the Seine, and end with air from the Department of the Rhone, or the Belgian frontier, according to the direction of the wind. So long as we had no analytical methods of sufficient delicacy to estimate with certainty the hundredth, or at least the tenth of a milligramme of carbonic acid, it was very difficult to determine the quantity in the air at a given time and place. It is frequently possible to analyze upon the plain air that has descended from the heights above, and to examine by bright daylight the effect of night upon the atmosphere. Still other difficulties show themselves in such investigations. It seems very easy to collect carbonic acid in potash tubes, and to determine its amount from the increase in weight of the tubes; but, alas! to how many sources of error is this method exposed. If the potash has been in contact with any organic substance, it will absorb oxygen. If the pumice that takes the place of the potash contains protoxide of iron, it will also absorb oxygen. In both cases the oxygen increases the weight of the carbonic acid.