fine powder, a duster wrapped round it torn to shreds, and a stout wooden box on which it was standing shattered. It is, however, perfectly safe to blow gases, for example, carbon dioxide, through the liquid or to shake it about in a narrow tube (3 mm. diameter). The gas has been caused to explode : (1.) By contact with organic matter, such as traces of impurity in sulphuric acid or phosphorus pentoxide. (2.) By scratching the tube containing it with a file. (3.) By shaking pieces of broken glass in it. The explosions, however, always stop at narrow tubes. The gas gradually decomposes over mercury, and is only slowly absorbed by caustic potash or soda; it will, in fact, bubble through solutions of these without much loss, and a moistened stick of caustic potash requires to be left for 10-12 hours in contact with the gas to effect complete absorption. Undistilled phosphorus pentoxide is acted upon, giving a red substance; it becomes hot and has even caused an explosion. Pure phosphorus pentoxide, however, is not affected; the gas can be kept over it for 2-3 days at 0° without much decomposition. The mixtures of chlorine peroxide and carbonyl sulphide were allowed to dry for 24 hours in the ice chest and then exploded either by means of a spark or by heating in an air-bath; the gases lost their deep green colour and became colourless. When the explosion was brought about by sparking, carbonyl chloride could always be detected among the products, its pungent odour being readily recognisable even in presence of chlorine and sulphur dioxide; if the explosion were caused by gradual heating, I could never clearly detect any of the chloride. In other respects, the results were identical. of carbon monoxide obtained were: The amounts No CO found about 2-4 per cent. of the COS is left as CO These results show that when carbon monoxide and oxygen are brought together, both being in the nascent condition and also heated by the flame of an explosion, combination is not complete. If longer drying of the chlorine peroxide were possible, no doubt a larger percentage of uncombined carbon monoxide would be found. IV. Inertness of pure Carbonyl Sulphide towards Oxygen. We have already seen that the further the purification of carbonyl sulphide is carried, the greater is the proportion left unburnt when mixtures of carbonyl sulphide with excess of oxygen are sparked. This can only be interpreted as showing that pure carbonyl sulphide will not explode on sparking with oxygen. So far, only one such sample has been obtained, and I have not found any method of purification which can be relied upon to give a sufficiently pure gas. In mixtures of carbonyl sulphide and nitrous oxide (which explode if not purified), after standing for some weeks over pure sulphuric acid, and for a further lengthy period over phosphorus pentoxide, an electric spark produces no explosion. On now decanting the gaseous mixture into another eudiometer, adding a drop of water, and passing a spark, a violent explosion takes place. Conclusions. The conclusions to be drawn from the above experiments are: 1. Pure carbonyl sulphide will not explode if sparked with oxygen. 2. If a small quantity of impurity is present, a flame traverses the whole tube on sparking the mixture, but combustion is not nearly complete; part of the carbonyl sulphide remains unburnt, and part is decomposed into carbon monoxide and sulphur, which likewise do not burn, although excess of oxygen and a small quantity of impurity are present. As, however, this quantity increases, combustion rapidly becomes more complete. 3. Mixtures of carbonyl sulphide and nitrous oxide require a larger quantity of impurity to cause combustion than mixtures of carbonyl sulphide and oxygen in the same circumstances. It seems probable that the quantity necessary for the latter mixture is also different from that required by mixtures of carbon monoxide and oxygen. 4. The state of affairs following on a violent reaction—such as explosion of carbon disulphide or chlorine monoxide, &c., has a very considerable influence in bringing about combination of carbon monoxide and oxygen. Whether this is a direct effect, or due to a heightening of the action of the "third substance," there is as yet little evidence. I can find no evidence that the nascent state, per se, is very effective. When the conditions were made as nearly as could be the same, there did not seem to be a great difference between the behaviour of nascent carbon monoxide and that of molecular carbon monoxide as studied by Dixon. Finally, I wish to thank Professor Dixon for much kindly help and advice given during the progress of this research. THE OWENS COLLEGE, MANCHESTER. XXXV.-Note on the Refraction and Magnetic Rotation of Hexamethylene, Chlorohexamethylene, and Dichlorohexamethylene. By SYDNEY YOUNG, D.Sc., F.R.S., and EMILY C. FORTEY, B.Sc. A SPECIMEN of hexamethylene was obtained by one of us (Trans., 1898, 73, 932) by the long-continued fractional distillation of Galician petroleum and was believed at the time to be pure. Its molecular refraction and magnetic rotation were determined by Dr. W. H. Perkin, sen., and the data obtained by him were published in the paper. It was afterwards found (Trans., 1899, 75, 873) that the hexamethylene could be partially, but not completely, frozen, and that it therefore contained a small quantity of another hydrocarbon, probably a heptane. By a series of fractional crystallisations, the hexamethylene was separated from the paraffin, and was finally found to melt practically contantly at +4.7°. Dr. Perkin has redetermined the refraction and magnetic rotation of this pure specimen, and has very kindly sent us the results for publication: The calulated value for the molecular rotation is 1·023 x 6 = 6·138; it will be seen, therefore, that the observed value differs from the calculated, being considerably lower. The refractive power is given in the following table: The temperature was 15° and the density d 15°/4° was 0·78224. Dr. Perkin has also redetermined the refraction and magnetic rotation of the specimens of monochlorohexamethylene and dichlorohexamethylene (Trans., 1898, 73, 932), because, although the rotation of the former was lower than he expected, and that of the latter was about correct, the refractive values of both were higher than the calculated by nearly a unit. He points out that, generally, if the refractive value is high, the magnetic rotation is proportionately a good deal higher. The new results, however, confirm those previously obtained. By an error, the calculated value for Cl displacing H in the molecular magnetic rotation was given in the paper referred to as 1.558 instead of 1.479 for the mono- and 1.391 for the di-displacement. The theoretical values for the two chlorine derivatives are therefore appended. The values obtained by Dr. Perkin are as follows: Cl disp. Has in monochloro-paraffins... 1479 The values for the dispersion Hy - Ha for monochlorohexamethylene, 1.622, and for dichlorohexamethylene, 2.180, differ but slightly from the corresponding values, 1.615 and 2.141, previously obtained. UNIVERSITY COLLEGE, BRISTOL. XXXVI.-Campholytic and Isolauronolic Acids. IN former papers (Walker, Trans., 1893, 63, 495; 1895, 67, 347), it was shown that when sodium ortho ethyl camphorate was electrolysed in aqueous solution, an unsaturated ethereal product was obtained at the anode in accordance with the following empirical equation: 2 CO2Et CH11 CO2Na + H2O = C ̧Н ̧ ̧° CO2Et + H2+ NaHCO3. This ethereal product, when fractionated and hydrolysed, yielded two isomeric unsaturated acids of the formula CH13 CO,H, one of which, campholytic acid, was a liquid, and the other, isolauronolic acid, a solid. These acids were also obtained by W. A. Noyes from B-camphoramic acid (Amer. Chem. J., 1894, 16, 505; 1895, 17, 421), and isolauronolic acid in particular has recently been the subject of numerous investigations, chiefly by Blanc and by W. H. Perkin, jun. By the action of alkalis on the dibromide of campholytic acid, a bromohydrocarbon is produced (Trans., 1893, 63, 502), a decomposition which Fittig's researches have shown to be characteristic of aß-dibromo-acids. Noyes found that the dibromide of isolauronolic acid undergoes a similar decomposition when neutralised with sodium carbonate, and was therefore led to regard the two unsaturated acids as stereoisomeric, a view which is strengthened by his discovery that campholytic acid is converted on standing with mineral acids into the isomeric isolauronolic acid. It is now generally admitted that isolauronolic acid is an unsaturated acid with the double bond in the aß-position with regard to the carboxyl group (Blanc, Ann. Phys. Chem., 1899, [vii], 18, 181; Perkin, Trans., 1898, 73, 810). In view of the fact that the relationship of campholytic and isolauronolic acids to camphoric acid is of the greatest importance in |