acetylene series, whilst samples of the gases escaping absorption were finally collected over water. Carbon was deposited in the decomposition tube, and at the end of one month 15 c.c. of liquid had been slowly condensed in the U-tubes, and in this liquid they detected benzene, naphthalene, anthracene, and some other aromatic hydrocarbons, present in quantities too small for determination. Faint traces only of precipitate were found in the ammoniacal cuprous chloride solution, whilst among the bodies absorbed by bromine they identified crotonylene tetrabromide, and the gas collected over water proved to be a mixture of methane and ethane. The absence of acetylene from the products obtained led them to the view that these products are formed directly by the action of heat upon ethylene. From the work of the earlier observers, the text-books have accepted the equation

1. C2H1 = С2+2H2

as representing the decomposition which takes place at a very high temperature, whilst, on the evidence of the work done by Marchand, and Buff and Hofmann, they represent the change taking place at a lower temperature by the equation

During my attempts to trace the actions taking place in the inner zone of luminous flames, I was struck by the complexity of the changes, and the absence of any evidence which would tend to confirm the second equation, and I have made the experiments detailed in the following paper in the hope of being able to trace the decomposition effected by heat in such simple hydrocarbons as ethylene, ethane, and methane.

The first step was to make experiments to corroborate the statement made by Day that, if ethylene is heated at 400° C. for a sufficient length of time, it is entirely decomposed, with formation of methane, ethane, and liquid products.

In order to do this an apparatus of the form used by Day in his experiments was employed, the only difference being that instead of using an air thermometer, the temperature was taken by means of a platinum and platinum-rhodium couple-as introduced by Le Châtelier, and fully described by Mr. C. Roberts-Austen-which had been previously carefully calibrated, employing salts of known fusing points. The ethylene was prepared for this and for all the following experiments by making a mixture of 25 parts by weight of rectified methylated spirit and 150 parts of strong sulphuric acid. The mixture was heated in a flask containing a layer of sand, and the gas evolved was washed by contact with strong sulphuric acid, and by passage through several bottles containing a strong solution of caustic soda. The gas was stored in a glass gas-holder for several days over water

containing sodic pyrogallate and sodic hydrate, to absorb all traces of oxygen, and on analysis gave

The gas was then passed through the combustion tube, until it was considered that all air had been displaced, and the tube was maintained for 100 hours at 400° C. Oil and a very small quantity of carbon deposited in the tube, and the volume decreased from 100 to 61. The gas was then removed from the tube and analysed.

In all the analyses described in this paper the following procedure was adopted. Carbon dioxide was absorbed by means of a 50 per cent. solution of sodic hydrate, the oxygen estimated by absorption with alkaline pyrogallate, the unsaturated hydrocarbons next absorbed by means of a solution of bromine in potassium bromide, care being taken to remove bromine vapour from the gas by agitation with caustic soda before measurement, the carbon monoxide was next estimated by acid cuprous chloride, and after removal of any acid fumes the residual gas was treated with paraffin oil, previously prepared for use by heating it over a water-bath for at least an hour.

Experiments show that in the case of mixtures of methane with higher members of the same group, agitation with paraffin prepared in this way, or mere standing in contact with it with occasional agitation for twenty to thirty minutes, will remove the ethane and any higher saturated hydrocarbons which may be present, together with a small proportion of the methane. The amount of residual methane can then be determined by explosion with oxygen, and subsequent estimation of the carbon dioxide formed, the volume of gas absorbed by the paraffin plus the volume resulting from explosion giving the total volume of saturated hydrocarbons. Details of the results obtained by this method of procedure when dealing with gaseous mixtures of known composition will be found, Jour. Soc. Chem. Industry,' vol. 10, p. 407.

The analysis of the heated ethylene gave

Carbon dioxide...

Oxygen....

6

0.82

0.00

the oxides of carbon and increase on the nitrogen showing that some air had remained in the tube.

A second analysis was now made, but instead of estimating the saturated hydrocarbons by first absorbing the higher members and some methane by paraffin, they were exploded with oxygen, and the methane calculated from the carbon dioxide amounted to 112.5 per cent., showing that ethane and probably even higher members of the series were present, results which fully bear out the statement made by Day as to the decomposition of the ethylene taking place at a temperature of 400° C., and also the statements made by Berthelot and by Day, that under the conditions of this experiment ethane or at any rate higher members of the CnH2n+2 series are formed as well as methane.

It would, however, be manifestly wrong to assume that the formation of the higher paraffins was a primary action, as keeping the hydrocarbons formed by the primary changes at a temperature of 400° C. might easily lead to the formation of secondary products by interaction between the gases.

It seemed much more probable that the character of the primary decompositions would be ascertained by rapidly heating the gas, and as rapidly removing the products of decomposition from the influence of heat, and that this would be effected by passing a regular current of the gas through a very narrow tube heated for 140 mm. to a known temperature.

The necessity for heating this tube to temperatures above 1000° C. practically limited the choice of material of which it could be made to fire-clay or platinum. It was at first feared that the use of the latter might interfere with the changes taking place, but a long series of comparative experiments, in which ethylene was decomposed by passing through (a) a pipe-stem glazed with borax, and (b) a platinum tube 2 mm. in diameter, both being heated to the same temperature, showed that the platinum tube was free from experimental objection unless a considerable percentage of oxygen was present, and that, even with a new tube, the decompositions were of the same nature as when the pipe-stem was employed.

Under these conditions the platinum tube possessed so many advantages over the clay pipe that in all the subsequent experiments a platinum tube, 2 mm. in diameter and about 40 cm. in length, was used, and, in order to accurately measure the temperature to which. the gas in the tube was heated, the following arrangement was devised :

Ethylene was stored in a gas-holder, and, after passing over calcic chloride to dry it, entered the platinum tube. In this tube the platinum and the platinum-rhodium thermo-couple was arranged in the following fashion :

The two wires are twisted together for a length of 3 mm., and the wires on either side of the twist are then passed through thin glass tubes, which are fused on to them; having been in this way coated with glass so that only the twist is exposed, they are passed through the platinum tube, the glass insulating the wire from the tube, and also keeping the thermo-junction in such a position that it registers the temperature of the gas in the tube, not that of the wall of the tube.

To each end of the platinum tube glass T-pieces are fitted, down the stems of which the wires pass to mercury seals; from the metal seals conducting wires lead to the resistance coils, the key, and a reflecting galvanometer. The products, after leaving the platinum tube, pass through a U-tube, cooled in ice and salt in order to condense any liquids, and then through a collecting tube, from which the sample of gas for analysis for gas is taken, thence to Volhard absorption flasks (containing ammoniacal silver nitrate for the estimation of acetylene), the flow of the gas through the apparatus being regulated by means of the aspirator bottle.

In the following set of experiments the ethylene, after purification from oxygen by standing over a dilute solution of sodic pyrogallate and sodic hydrate, was passed through the tube at the rate of about 10 c.c. per minute.

These experiments are of considerable interest, as they throw some light upon the changes taking place during the heating of ethylene.

The gas being passed through 140 mm. of heated tube, no change takes place until a temperature of 800° C. is reached, when traces of acetylene are observed; between 800° and 900° C. the acetylene increases in quantity, and large quantities of methane are generated, accompanied by liquid products. This action increases until just below 1200° C., when hydrogen begins to appear amongst the products of decomposition, whilst the moment the liberation of hydrogen commences, carbon also is deposited; and the formation of oil decreases until close upon 1500° C., when the decomposition of the ethylene is practically complete, and the products of decomposition are mainly hydrogen with some undecomposed methane, and a copious deposit of carbon.

In each experiment the products of decomposition were examined to see if any member of the CnH2n+2 group other than methane was present, and in no case could any be detected. This seems to point strongly to the conclusion that the ethane formed in the previous experiment, in accordance with the experiments of Day and Berthelot, had its origin as a secondary, and not as a primary, product of decomposition, but it was clear that to determine this point other experiments must be made, to see if under these conditions of tem

Table I.-The Action of Heat upon flowing Ethylene.

Percentage of ethylene in _original gas.

Temperature of gas in decomposing tube

96.78 96.78 96.78 94.8

94.8 98.91 98.91

600° C. 700° C. 800° C. 900° C. 1000° C. 1200° C. 1500° C.

perature a slower rate of flow was attended by generation of any saturated hydrocarbons higher than methane.

In order to do this, the same arrangement of apparatus was employed, but the rate of flow of ethylene was diminished to 4-2 c.c. per minute, the gas, after heating, being collected as before, and an analysis made in which the saturated hydrocarbons were calculated from the results of the paraffin absorption and subsequent explosion. A second analysis was then made of the heated gas without any paraffin absorption, and the volume of carbon dioxide formed on explosion calculated as methane. If the volume so obtained is found to be greater than that originally given by the combined paraffin absorption and subsequent explosion, it is held to be evidence that some higher member of the paraffin series must be present. In order to ascertain if any unsaturated hydrocarbon other than ethylene is present in the gas after heating, a sample of the gas is exploded with oxygen, and the carbon dioxide estimated; should the amount formed exceed the carbon dioxide calculated from the second analysis, it may be taken as evidence that an unsaturated hydrocarbon higher than ethylene is present.