232 



NATURE 



[July 7, 1892 



crystals. The vapour of nickel carbonyl possesses a character- 

 istic odour and is poisonous, but not more so than carbonic 

 oxide gas. Prof. McKendrick has studied the physiological 

 action of this liquid, and has found that, when injected sub- 

 cutaneously in extremely small doses in rabbits, it produces an 

 extraordinary reduction in temperature, in some cases as much 

 as 12°. 



The liquid can becompletely distilled withoutdecomposition, but 

 from its solution in liquids of a higher boiling-point it cannot be 

 obtained by rectification. On heating such a solution the com- 

 pound is decomposed, nickel being separated in the liquid, while 

 carbonic oxide gas escapes. I will try to demonstrate this by 

 an experiment. 



We have here a solution of the substance in heavy petroleum 

 oil, which you will, in a few minutes, see turns completely 

 black on heating by the separation of nickel, while a gas 

 escapes which is carbonic oxide. 



In a similar way, when the nickel carbonyl is attacked by 

 oxidizing agents, such as nitric acid, chlorine, or bromine, it is 

 readily broken up, nickel salts being formed, and carbonic 

 oxide being liberated. Sulphur acts in a similar way. Metals, 

 even potassium, alkalies, and acids, which have no oxidizing 

 power, will not act upon the liquid at all, nor do the salts of 

 other metals react upon it. The substance behaves therefore, 

 chemically, in an entirely different manner from potassium car- 

 bonyl, and does not lead, as the other does, by easy methods to 

 complicated organic compounds. It does not show any one of 

 the reactions which are so characteristic for organic bodies con- 

 taining carbonyl, such as the ketones and quinones ; and we 

 have not been able, in spite of very numerous experiments, 

 either to substitute the carbonic oxide in this compound by other 

 bivalent groups, or to introduce the carbonic oxide by means of 

 this compound into organic substances. 



By exposing the liquid to atmospheric air, a precipitate 

 of carbonate of nickel is slowly formed of varying composi- 

 tion, which is yellowish-white if perfectly dry air is used, and 

 varies from a light green to a brownish colour if more or 

 less moisture is present. We have found all these precipitates 

 to dissolve easily and completely in dilute acid, with evolution 

 of carbonic acid, leaving ordinary nickel salts behind, and can 

 therefore not agree with the view propounded by Prof. ]3erthelot, 

 in a communication to the French Academy of Sciences, that 

 these precipitates contain a compound of nickel with carbon and 

 oxygen, comparable to the so-called oxides of organo-metallic 

 compounds. In the same paper Prof. Berthelot has described a 

 beautiful reaction of nickel carbonyl with nitric oxide, which 

 we will now show you. You will notice the intense blue 

 coloration which the liquid solution of nickel carbonyl in 

 alcohol assumes by passing the nitric oxide through it. Prof. 

 Berthelot has reserved to himself the study of this interesting 

 body, but has so far not published anything further about it. 



The chemical properties of the compound I have just de- 

 scribed to you are without parallel ; we do not know a single 

 substance of similar properties. It became, therefore, of special 

 interest to study the physical properties of the compound. 



Prof. Quincke, of Heidelberg, has kindly determined its 

 magnetic properties, and found that it possesses in a high 

 degree the property discovered by Faraday, and called by him 

 diamagnetism, which is the more remarkable, as all the other 

 nickel compounds are paramagnetic. He also found that it is 

 an almost perfect non-conductor of electricity, in this respect 

 differing from all other nickel compounds. 



The absorption spectrum, and also the flame spectrum, of our 

 compound are at present under investigation by those inde- 

 fatigable spectroscopists, Profs. Dewar and Liveing, by whose 

 kindness I am enabled to bring before you, in advance of a 

 paper they are sending to the Royal Society, some of the 

 interesting results they have obtained. We have here a photo- 

 graph of the absorption spectrum, obtained by means of a hollow 

 prism through quartz plates filled with nickel carbonyl, through 

 which the spark spectrum of iron is passed, which is photo- 

 graphed on the same plate. You see that the whole of the ultra- 

 violet rays of the iron spectrum have disappeared, being com- 

 pletely absorbed by the nickel carbonyl, which is thus quite 

 opaque for all the rays beyond the wave-length 3820. The 

 spectrum of the highly luminous flame of nickel carbonyl, 

 which I have shown you before, is quite continuous ; but if the 

 nickel carbonyl is diluted with hydrogen, and the mixture burnt 

 by means of oxygen, the gases burn with a bright yellowish- 

 green flame without visible smoke ; and the spectrum of this 



NO. II 84, VOL. 46] 



flame shows in its visible part, on a background of a continuous 

 spectrum, a large number of bands, brightest in the green, but 

 extending on the red side beyond the red line of lithium, and 

 on the violet side well into the blue. These bands cannot be 

 seen on the photograph which I will now show you, the visible 

 part of the spectrum appearing continuous ; but beyond the 

 visible part the photograph shows a large number — over fifty— 

 of well-defined lines in the ultra-violet, I will show you these 

 lines in another photograph taken with greater dispersion, and 

 on which has also been photographed the spark spectrum of 

 nickel. You will see that all these lines correspond absolutely 

 to lines appertaining to the spark spectrum ; in fact, the greater 

 part of the lines in the spark spectrum are also shown in this 

 flame spectrum. We have here another and very striking 

 example of the fact discovered on the same day by Profs. Dewar 

 and Liveing, and by Dr. Huggins, that 'the spectrum of 

 luminous flames is not always continuous throughout its whole 

 range, a fact which was at one time much debated and dis- 

 cussed. 



One of the most remarkable discoveries made within the 

 precincts of this Institution by that illustrious man whose 

 centenary we celebrated last year was that of the connection 

 between magnetism and light, which manifests itself when a 

 beam of polarized light is sent through a substance while it is 

 subjected to a strong magnetic field, under whose influence the 

 beam of light is rotated through a certain angle. Dr. W. H. 

 Perkin has prosecuted this discovery of Faraday's by a long 

 series of most elaborate researches, and has established the fact 

 that this power of magnetic rotation of various bodies has a 

 definite relation to their chemical constitution, and enables us to 

 gain a better insight into the structure of chemical compounds. 

 Dr, Perkin has been good enough to investigate the power of 

 magnetic rotation of the nickel carbonyl, and has found it quite 

 as unusual as its chemical properties, and to be, with the sole 

 exception of phosphorus, greater than that of any other sub- 

 stance he has yet examined. 



The power of different bodies of refracting and dispersing a 

 ray of light has been shown by the beautiful and elaborate re- 

 searches undertaken many years ago by Dr. Gladstone — who has 

 given an account of them in this theatre in 1875, and who has 

 since continued them with indefatigable zeal — -to throw a con- 

 siderable light upon the constitution of chemical compounds. 



I have investigated the refractive and dispersive powers of 

 nickel carbonyl in Rome, in conjunction with Prof. Nasini. We 

 found that the atomic refraction of nickel in the substance is 

 nearly two and a half times as large as it is in any other nickel 

 compound — -a difTerence very much greater than had ever before 

 been observed in the atomic refraction of any element. To give 

 you some idea how these figures are obtained, Mr. Lennox will 

 now throw on to the screen a beam of light through two super- 

 posed prisms, one filled with nickel carbonyl and the other with 

 alcohol. You will notice that the lines of the spectrum on the 

 top are turned much further to the left, showing the nickel 

 carbonyl to possess a much greater power of refraction than 

 alcohol, and you will also notice that it is much wider than the 

 bottom spectrum, which shows the greater dispersive power of 

 the nickel carbonyl. 



It is now generally supposed that, if one element shows different 

 atomic refractive powers in different compounds, it enters with a 

 larger number of valencies into the compound which shows a 

 higher refractive power. In accordance with this view, the 

 very much greater refractive power of the nickel in the carbonyl 

 would find an explanation in assuming that this element, which 

 in all its other known combinations is distinctly bivalent, exer- 

 cises in the carbonyl the limit of its valency, viz. 8, assigned 

 to it by Mendeleeff", who placed it in the eighth group in his 

 Table of Elements. This would mean that the one atom of 

 nickel contained in the nickel carbonyl is combined directly with 

 each of the four bivalent atoms of carbonyl, each of which would 

 saturate two of the eight valencies of nickel, as is shown by this 

 formula — 



O 



C 

 il 

 O : C = Ni = C : O 

 II 

 C 



