I shall now give you a general idea of the apparatus. Into a common glass flask of carafe shape, and flat-bottomed, of 30 cubic inches capacity, are passed two small tubes, at the end of one of which is attached a small square silver table, 1 inch in length. A little water having been inserted, the flask is inverted, and the table is placed exactly 1 inch from the inverted bottom, so that the contents of air right above the table are 1 cubic inch. This observing table is divided into 100 equal squares, and is highly polished, with the burnishing all in one direction, so that during the observations it appears dark, when the fine mist-particles glisten opal-like with the reflected light in order that they may be more easily counted. The tube to which the silver table is attached is connected with two stop-cocks, one of which can admit a small measured portion of the air to be examined. The other tube in the flask is connected with an air-pump of 10 cubic inches capacity. Over the flask is placed a covering, coloured black in the inside. In the top of this cover is inserted a powerful magnifying glass, through which the particles on the silver table can be easily counted. A little to the side of this magnifier is an opening in the cover, through which light is concentrated on the table. To perform the experiment, the air in the flask is exhausted by the air-pump. The flask is then filled with filtered air. One-tenth of a cubic inch of the air to be examined is then introduced into the flask, and mixed with the 30 cubic inches of dustless air. After one stroke of the air-pump, this mixed air is made to occupy an additional space of 10 cubic inches; and this rarefying of the air so chills it that condensation of the water-vapour takes place on the dust-particles. The observer, looking through the magnifying-glass upon the silver table, sees the mist-particles fall like an opal shower on the table. He counts the number on a single square in two or three places, striking an average in his mind. Suppose the average number upon a single square were five, then on the whole table there would be 500; and these 500 particles of dust are those which floated invisibly in the cubic inch of mixed air right above the table. But, as there are 40 cubic inches of mixed air in the flask and barrel, the number of dust-particles in the whole is 20,000. That is, there are 20,000 dust-particles in the same quantity of common air (one-tenth of a cubic inch) which was introduced for examination. In other words, a cubic inch of the air contained 200,000 dust-particles--nearly a quarter of a million. The day I used the instrument we counted 4,000,000 of dust-particles in a cubic inch of the air outside of the room, due to the quantity of smoke from the passing trains. Dr. Aitken has counted in 1 cubic inch of air immediately above a Bunsen flame the fabulous number of 489,000,000 of dust-particles.