'If we had time to examine them well,' continued Pasteur, 'we should see them, though agitated with various movements, falling downwards more or less quickly. It is thus that all objects become covered with dust--the furniture, the table, the mercury in this basin. Since this mercury was taken from the mine, how much dust must have fallen upon it, to say nothing of all that has been intimately mixed up with it during the numerous manipulations to which it has been subjected in the laboratory? It is not possible to touch this mercury, to place the hand in it, or a bottle, without introducing into the interior of the basin the dust which lies on its surface. You will now see what takes place.' Projecting, in the darkness, the beam of light upon the basin of mercury, the liquid metal shone forth with its usual brilliancy. Pasteur then sprinkled some dust upon the mercury, and, plunging a glass rod into it, the dust was seen to travel towards the spot where the rod entered the mercury, and to penetrate into the space between the glass and the metal. 'Yes,' exclaimed Pasteur with a voice which gave evidence of the sincerity of his conviction, 'yes, M. Pouchet had removed the germs from the water and from the hay, but he had neglected to remove the dust from the surface of the mercury. This is the cause of his error; this is what has vitiated his whole arrangement.' Pasteur then instituted experiments exactly similar to those of Pouchet, but taking care to remove every cause of error which had escaped the latter. He employed a glass bulb with a long neck, which he bent, and connected with a tube of platinum placed in a furnace, so that it could be heated nearly to redness. In the bulb he placed some very putrescible liquids--urine for example. When the furnace which surrounded the platinum tube was in action, Pasteur boiled the liquid for some minutes, then he allowed it to cool, keeping the fire around the platinum tube still active. During the cooling of the bulb the external air was introduced, after having first travelled through the red-hot platinum tube. The liquid was thus placed in contact with air whose suspended germs were all burnt up. In an experiment thus carried out, the urine remains unchanged--it undergoes only a very slight oxidation, which darkens its colour a little--but it exhibits no kind of putrefaction. If it be desired to repeat this experiment with alkaline liquids, such as milk, the temperature must be raised a little above the boiling point--a condition easily realised with the apparatus just described. It is only necessary to connect with the free extremity of the platinum tube a glass tube bent at right angles, and to plunge the latter to a depth of some centimeters into a basin of mercury. In these circumstances ebullition goes on under a pressure greater than that of the atmosphere, consequently at a temperature higher than 100 degrees Centigrade.