Makcn 14, 1907] 
ative ions in flames are free electrons, and not atoms 
oaded corpuscles. 
iteresting results, suggesting a field for further investi- 
on, were found for the gradient when salt was 
orised beneath both electrodes. 
anuary 24.—‘ On a New Iron Carbonyl and on the 
ion of Light and Heat on the Iron Carbonyls.” By 
James Dewar, I.R.S., and Dr. H. O. Jones. 
‘he paper contains an account of the extension of the 
_xperiments, previously described, on the action of light 
on the liquid iron pentacarbonyl and on the action of 
heat on the resulting solid compound, diferrononacarbonyl. 
In the course of the experiments, new and interesting 
‘observations were made, and a new compound of iron and 
carbon monoxide discovered. 
The action of light on iron pentacarbonyl alone or in 
solution results in the formation of Fe,(CO), and carbon 
monoxide, except in two cases, (1) when the solvent is 
nickel carbonyl, and (2) when the temperature is above 
SGraG. 
The absence of any action due to light above 56° C. 
has been confirmed by using tubes fitted with a small 
manometer, which showed no change of pressure when 
no solid was deposited, and showed that the deposition of 
solid was a delicate test for any action. 
At 35° C. solid was deposited, and pressure developed 
in five minutes in sunlight. 
At 45° C. to 50° C. solid was deposited, and pressure 
developed in thirty minutes in sunlight. 
At 56° C. no solid was deposited, and no pressure 
developed in five to twelve hours in sunlight. 
The velocity of the reaction was measured, and it was 
found to be a reaction of the ‘‘ first order.’’ The rate of 
decomposition was compared with the rate of the reaction 
between ferric chloride and oxalic acid, which was in- 
vestigated by Lemoine; it was found that iron carbonyl 
was slightly more sensitive to light than the mixture used 
by Lemoine. 
The rate of the reverse action of carbon monoxide on 
the solid, which takes place in the dark, was also 
measured, and was found to be very small at the ordinary 
temperature, but to have a fairly normal temperature 
coefficient; the velocity was approximately trebled for an 
increase of 10° C. 
Thus a reasonable explanation of the absence of any 
action of light above 56° C. can be suggested. If the 
direct action induced by light has a very small tempera- 
ture coefficient, as the reaction investigated by Lemoine 
thas, the reverse action being about 240 times as rapid at 
56° C. as at 16° C. would easily prevent the accumula- 
tion of appreciable quantities of the products of the de- 
composition. 
The action of heat on diferrononacarbonyl alone has 
been shown to produce a decomposition represented by 
the equation 2Fe,(CO),=3Fe(CO),+Fe+3CO. When the 
solid was heated in the presence of hydrocarbons, ether, 
‘bromobenzene, or iron pentacarbonyl, however, green solu- 
tions were obtained, iron pentacarbonyl was produced, but 
no gas was evolved; if alcohol, pyridine, acetone, or 
acetonitrile was the liquid used, red solutions were 
obtained, but no gas was evolved; with nickel carbonyl 
as solvent, gas was evolved and iron deposited. From the 
green solutions, under suitable conditions, lustrous green 
crystals were deposited; these were found to be a new 
compound, iron tetracarbonyl, x{Fe(CO),}, where x is 
large, probably about 20. 
The new compound has a molecular volume of 84 for the 
unit Fe(CO),, and is very stable. It is not decomposed 
by hot concentrated hydrochloric acid, is attacked by hot 
concentrated sulphuric acid giving carbon monoxide and 
ferrous sulphate, and is readily decomposed by cold nitric 
acid. 
Iron tetracarbonyl dissolves in hydrocarbons, ether, iron 
pentacarbonyl, nickel carbonyl, acetone, and acetonitrile 
to give green solutions which do not change on heating 
for a short time, and which deposit the green compound 
unchanged when evaporated out of contact with air; in 
pyridine and alcohol the compound dissolves to give a 
NO. 1950. VOL. 75] 
NATURE 
477 
green solution, which changes slowly in the cold and 
rapidly on heating into a red solution. 
The green solutions exhibit a characteristic absorption 
band in the yellow, while the red 
selective absorption. 
solutions show no 
Chemical Society, February 21.—Prof. k. Meldola, F.R.S., 
president, in the chair.— The constitution of hydroxyazo- 
compounds: W. B. Tuck. Several of the hydroxyazo- 
compounds were- examined spectrographically, and it was 
found that the absorption spectra of the p-hydroxy- 
compounds agree closely with those of their derivatives. 
The ethers of o-compounds also agree with the p-com- 
pounds, but the benzoyl derivatives are similar to benzo- 
quinonebenzoylphenylhydrazone.—The influence of solvents 
on the rotation of optically active compounds, part ix., a 
new general method for studying intramolecular change : 
T. S. Patterson and A. McMillan.—Displacement of 
halogens by hydroxyl, i., the hydrolytic decomposition of 
hydrogen and sodium monochloroacetates by water and by 
alkali, and the influence of neutral salts on the reaction 
velocities: G. Senter. On the basis of the results 
obtained, the hypothesis put forward by R. J. Caldwell, 
that the accelerating influence of neutral salts on certain 
catalytic actions is due to the withdrawal of water and 
consequent concentration of the reacting substances, was 
criticised. It is considered that the effect in question is 
due to the action of the ions of the salt on H’ and OH’ 
ions.—The interaction of ammonium salts and the con- 
stituents of the soil: A. D. Hatt and C. T. Gimingham. 
—The reduction products of o- and p-dimethoxybenzoin : 
J. C. Irvine and Miss A. M. Moodie.—Constituents of 
natural indigo, part ii.: A. G. Perkin. Numerous Java 
indigos have been found to contain kzmpferol. The 
leaves of Indigofera sumatrana contain a_ trace of what 
is probably kampferol.—The velocity of hydrolysis of 
aliphatic amides: J. C. Crocker.—The rates of reaction 
of formamide, acetamide, propionamide, butyramide, 1so- 
butyramide, valeramide, capronamide with hydrochloric 
acid, have been determined at from 40° to 80°. The re- 
actions are bimolecular, and the order of the relative 
reactivities is the same for each temperature. A relation 
between the reactivity of the amides at constant tempera- 
ture and the strength of the corresponding organic acids 
was indicated.—The rusting of iron: W. R. Dunstan. 
In order to explain the fact that the rusting of iron can 
take place in the absence of carbonic acid, and that only 
iron, oxygen, and liquid water are necessary, the work- 
ing hypothesis was suggested that the formation of 
hydrogen peroxide is concerned in the change, rusting 
being prevented by those substances which are capable 
of decomposing the peroxide (Trans. Chem. Soc., 1905, 
Ixxxvii., 1548). The results of further experiments carried 
out by the author show that iron rusts freely in 
the absence of carbonic acid, provided that iron, 
oxygen, and liquid water are brought together.—Contri- 
butions to the chemistry of the rare earths, part ii 
M. Esposito. The methods of Muthmann and Bohm and 
Pattinson and Clarke for the preparation of ceria have 
been found to give a fairly pure product. Lanthana and 
old didymia can be separated by fractional crystallisation 
of the oxalates from strong nitric acid. Lanthana is best 
obtained by fractional crystallisation of the double 
ammonium nitrates. After one hundred and ten fraction- 
ations, lanthana, praseodymia, and neodymia were 
obtained in a state of considerable purity.—Derivatives of 
multivalent iodine, part iii., action of heat on iodobenzene 
dichloride and on the m- and p-nitro- and p-chloro- 
derivatives: W. Caldwell and E. A. Werner.— he 
organic phosphorus compound formed by yeast-juice from 
soluble phosphates. Preliminary notice: W. J. Young. 
A lead salt of the compound was prepared from the 
fermentation mixture by first removing any free phosphate 
by magnesium nitrate, and then adding lead nitrate. 
Analyses of two preparations, in which the carbon, 
hydrogen, lead, and phosphorus were determined, gave 
the empirical formula C,H,O,PPb. From a solution of 
this the free acid can be obtained, which reduces Fehling’s 
solution, gives Mohlisch’s a-naphthol reaction, is slightly 
dextrorotatory, and can be titrated with allalis.—Experi- 
