[Footnote: Read before the British Association, Southampton Meeting, Section B, 1882.]

By HAROLD B. DIXON, M.A., Millard Lecturer in Chemistry, Balliol and Trinity Colleges, Oxford.

Two years ago I had the honor of showing before the Chemical Section of the British Association some experiments, in which a well-dried mixture of carbonic oxide and oxygen was submitted to electric sparks without exploding.[1] It was further shown that the introduction of a very minute quantity of aqueous vapor into the non-explosive mixture was sufficient to cause explosive combination between the gases when the spark was passed. The hypothesis advanced to account for the observed facts was that carbonic oxide does not unite directly with oxygen at a high temperature, but only indirectly through the intervention of water-vapor present, a molecule of water being decomposed by one of carbonic oxide to form a molecule of carbonic acid and one of free hydrogen, and the latter uniting with the oxygen to re-form a molecule of water, which again undergoes the same cycle of changes, till all the oxygen is transferred to the carbonic oxide:

HO + CO = H + CO

H + O = HO

[Footnote 1: "Report of British Association," 1880, p. 503.]

For such a series of reactions a comparatively few molecules of water would suffice, and the change produced by their alternate reduction and oxidation would come under the old term of "catalytic action," inasmuch as the few water molecules present at the beginning are found in the same state at the completion of the reaction.

The truth of this hypothesis has since been confirmed by experiments I have made on the incomplete combustion of mixtures of carbonic oxide and hydrogen; and on the velocity of explosion of carbonic oxide and oxygen with varying proportions of aqueous vapor. I therefore thought a description of the more convenient methods lately devised as lecture experiments for showing the influence of water on the combustion of carbonic oxide would not be uninteresting to the Section.

A glass tube from 18 inches to 2 feet long, closed at one end, and provided with platinum wires, is bent near its open end so that the shorter arm makes an angle of about 60° with the longer arm. The tube, held by a clamp, is heated in a Bunsen flame, and is then filled with mercury heated to about 130° C. The mixture of gases is then made to displace a portion of the mercury by forcing it through a fine tube, which is connected by a steel cap to the eudiometer of McLeod's gas apparatus, and passes down through the mercury in the shorter arm of the experimental tube. When a sufficient quantity of the gaseous mixture has been collected in the longer arm, some dry phosphoric oxide is introduced in the following way: A small glass tube is heated, packed with the dry powder, and pushed down into the shorter arm of the experimental tube. With a hot glass rod the phosphoric oxide is pushed out at the bottom of the small tube, and passes up into the gaseous mixture in the longer arm. After standing for a few hours in contact with the phosphoric oxide, the gases may be submitted to strong sparks from a Leyden jar without igniting. Care must be taken that none of the oxide comes in contact with the platinum wires, for if any sticks to the wires it becomes heated by the passage of the sparks, and gives off enough water to determine the explosion. In this way I have prepared several specimens of a non-explosive mixture of carbonic oxide and oxygen in the proper proportions to form carbonic acid. Some of these tubes have been submitted without explosion to sparks from a large Leyden jar, to a continuous succession of sparks from a Holtz machine, and to the discharge of a Ruhmkorff's coil, that heated the platinum wires between which it passed to bright redness. Other tubes which withstood the passage of the sparks from a Leyden jar, when submitted to the discharge of the coil, exploded after a few seconds when the platinum wires became red-hot. This I think may probably be attributed to hydrogen, occluded by the platinum, being given off on heating, and forming steam with the oxygen present.

For an easy and striking lecture experiment, I employ a tube open at both ends and bent like a W. The two open arms are short and the platinum wires are fixed at the highest bend. The tube is filled with hot mercury--one of the ends being closed by a caoutchouc stopper for the purpose--and a dry mixture of 5 volumes of air and 2 volumes of carbonic oxide is introduced into the bent tube over the mercury. A little phosphoric oxide is passed up one arm. After a few minutes the gases may be submitted to the spark without exploding. A little water may then be introduced through a pipette into the other arm; and if the spark is passed directly the gases ignite in the wet and not in the dry arm of the tube.

The admixture of the inert nitrogen renders a larger quantity of aqueous vapor necessary for the explosion than when only carbonic oxide and oxygen in proper proportion are present.