however, somewhat more stable than the plumbic salts, and the nitrate, sulphate, and oxalate have been prepared.

Like lead dioxide, manganic oxide, Mn2O3, is not attacked by acetic acid, the oxide Mn,O,, however, dissolves readily enough, and on adding a little water, crystals of Mn(C,H,O2)3,2H2O are deposited (Christensen, J. pr. Chem., 1883, [2], 28, 1). This salt is apparently the most stable of the manganic compounds, and its behaviour towards reagents is similar to that of lead tetracetate; a sulphate, Mng(SO) H2SO,,8H2O; phosphates, MnPO,,H,O, and MnNaP207,5H2O, and an arsenate, MnAsO,,H2O, have been prepared from it. Christensen has also described a fluoride, M¤F,,3H2O, and numerous double salts of the type 2M'F,MnF3 (J. pr. Chem., 1887, [2], 35, 57).

Cobaltic acetate exists in the solution prepared by dissolving hydrated cobaltic oxide in strong acetic acid (Beetz, Ann. Phys. Chem., 1844, 61, 472). With the exception of the sulphate described by Marshall (Trans., 1891, 59, 767), it seems to be the most stable of the simple cobaltic salts. When the oxide, CrO3, dissolves in glacial acetic acid, an acetate may possibly be found; we have not, however, been able as yet to obtain any evidence of this.*

List of Salts of Quadrivalent Lead.

Leaving out of consideration the numerous organo-metallic derivatives of lead, and omitting a few compounds whose nature is not yet fully elucidated (Wells, Zeit. anorg. Chem., 1895, 9, 305), the following list comprises the salts of quadrivalent lead which have, up to the present, been described by Brauner, Classen and Zahorski, Friedrich, Goebbels, Wells, and ourselves.

Pb(SO1),?

2PbC1,5(C,NH, HCI) (pyridine salt).

2PbC1,5(C,NH,,HCI) (lutidine salt).

3PbC1,7(C,NH,,HCI) (picoline salt).

РЬ(НРО,),? Pb(C2H2O2) Pb(C3H2O2)
Pb(HASO4)2.

*The remarkable compound obtained by Tafel (Ber., 1894, 27, 816) by acting with Na2O, on acetic acid, does not appear to belong to this class of substances.

VOL. LXIX.

R

Brauner, Friedrich, and Wells have pointed out that several of these compounds present analogies not only in composition, but also in behaviour and crystalline form, with certain of the stannic salts, and the relationship between lead and tin is further exhibited by the existence of plumbates allied to the stannates. Of them the potassium, sodium, barium, strontium, and calcium salts have been prepared, whilst the two oxides Pb2O, and Pb,O,, may be regarded as derivatives of meta- and ortho-plumbic acids respectively, and may

therefore be written PbQ,PbO, and

РЬО
РЬО

In conclusion, the principal results of our work, and our deductions from them, may be summed up as follows.

1. The substance obtained by dissolved Pb,O, in glacial acetic acid has the molecular formula Pb(C,H,O2). It is decomposed quantitatively by water into lead dioxide and acetic acid; hydrochloric acid converts it into lead tetrachloride, and orthophosphoric acid into a phosphate. Its general behaviour is analogous to that of the thallic and manganic acetates, and its molecular volume agrees with that calculated for the acetates by Schröder's empirical law.

These facts lead us to regard it as a salt of lead dioxide.

2. A similar propionate exists.

3. When acted on by orthophosphoric acid, the tetracetate is converted into a phosphate, to which in all probability the formula Pb(HPO), must be assigned.

4. Numerous lead salts exist which bear the same relation to the stannic salts that the ordinary lead compounds do to the stannous salts. The former should therefore be termed plumbic salts, and the latter plumbous salts.

5. Like stannic oxide, lead dioxide is capable of playing the part of either an acid or a basic oxide. The appropriate name for it is plumbic oxide, which should be used in preference to the term peroxide, which in this connection is both unsuitable and misleading.

Mineralogical Museum,

Cambridge.

XXIV. The Acetylene Theory of Luminosity. By VIVIAN B. LEWES, Royal Naval College, Greenwich. EARLY in 1892 (Trans., 1892, 61, 322), I read a paper before the Chemical Society, in which I showed that in a luminous hydrocarbon flame the baking effect of the outer zone of intense combustion converted a very large proportion of the unsaturated hydrocarbons

present in the inner zone into acetylene, the maximum production of this compound taking place just before the commencement of luminosity.

In a second paper, communicated to the Royal Society in 1895 (Proc. Roy. Soc., 57, 450), I gave experimental reasons for considering that the acetylene so formed on its decomposition by heat into carbon and hydrogen was the main factor in producing luminosity, as the heat developed by this decomposition raised the carbon particles to a temperature above that of the flame.

At the time that the latter paper was published, I was unaware that any other work had been done in this direction, but I find that Professors Dewar and Liveing, in their beautiful work on the spectra of carbon and its compounds, had to a great extent forestalled the conclusions to which I had arrived from a totally different standpoint.

In their paper "On the Origin of the Hydrocarbon Flame Spectrum" (Proc. Roy. Soc., 1882, 34, 427), they say, when speaking of the flame of cyanogen and acetylene, "Both of these compounds decompose with evolution of heat, in fact they are explosive compounds, and the latent energy in the respective bodies is so great that, if kinetic in the separated constituents, it would raise the temperature between 3000° and 4000°. The flames of cyanogen and acetylene are peculiar in respect that the temperature of individual decomposing molecules is not dependent entirely on the temperature generated by the combustion, which is a function of the tension of dissociation of the oxidised products, carbonic acid and water. We have no means of defining with any accuracy the temperature which the particles of such a flame may reach. We know, however, that the mean temperature of the flames of carbonic oxide and hydrogen lies between 2000° and 3000°, and if to this be added that which can be reached independently by the mere decomposition of cyanogen or acetylene, then we may safely infer that the temperature of individual molecules of carbon, nitrogen, and hydrogen in the respective flames of cyanogen and acetylene, may reach a temperature of from 6000° to 7000°.

"A previous estimate of the temperature of the positive pole in the electric arc made by one of us gave something like the same value.

*

"The formation of acetylene in ordinary combustion seems to be the agent through which a very high local temperature is produced

I extremely regret not having known of this most valuable work before, and take this, the first opportunity afforded me, of drawing attention to it.

I also find that M. Guéguen (Compte rendu du Société Technique de l'Industrie du Gaz au France, 1884, 142) pointed out that it was very probable that luminosity in hydrocarbon flames is due exclusively to the production of rays furnished by the molecules of gas highly heated by chemical changes, and that it must be borne in mind that the heating from exterior sources would not suffice whatever its power. He also draws attention to the fact that luminous combustion is caused by bodies which are endothermic, and from which heat is liberated during decomposition.

In November, 1895, Professor Smithells read a paper before the Chemical Society in which he criticises some of the less important points brought forward by me in my previous papers, and without attempting to disprove the fact that acetylene undergoes luminous decomposition when heated apart from air or oxygen, the principal fact upon which the acetylene theory of the luminosity of hydrocarbon flames is based, comes to the final conclusion that if the criticism the offers is just "then the acetylene theory of luminosity will share the fate of the dense hydrocarbon' theory."

The first portion of his paper is devoted to the measurement of flame temperature by means of the Le Chatelier thermo-couple, and he comes to the conclusion that the experiments show two things. 1. The fallacious results that may be obtained by not disposing the couple with due regard to the conformation of the zone of flame to be measured. 2. The difficulty of ascertaining the increase of temperature contributed by the chemical changes occurring in any one ..spot.

The first conclusion is self-evident, and the second I have already warmly endorsed in the paper which Professor Smithells is criticising (Proc. Roy. Soc., 1895, 57, 452).

He then proceeds to explore the temperatures existing in a luminous flat flame of coal gas, using for this purpose a No. 4 Bray's union jet burner at a gas pressure of 21 in. of water. As Professor Smithells considers the temperature determinations which he made as practically valueless, it is not necessary to criticise the pressure used, which should have been 7/10ths rather than 2 in. Considering, however, the abnormal character of the flame with which he was dealing, his figures accord fairly well with those I have given as representing the gradual rise of temperature in the inner zone of a luminous flat flame, and this is of interest, as the methods employed were slightly different, and the coincidence of result goes some way towards establishing the probability of the figures.

In considering the temperature in the flat flame, Professor Smithells says: "To obtain any useful measurements for this flame it is obvious that it can be explored in one way only, namely, by lay

ing the junction and the adjacent wires in a horizontal straight line, and introducing it along the flat face of the flame so that it may be placed with a considerable length of wires immersed symmetrically in any one sheath and passing through parts of the flame in like condition." Having then taken the temperature of the outer and inner zones in this way, he proceeds to say: "The measurements for the inner parts of the flame above given have only a negative significance. It is obvious that they cannot be true temperatures as the wires have always to be thrust through the hot outer sheath of flame, and, moreover, the couple would attain by radiation from the outer walls a higher temperature than the rapid stream of gas in which it is immersed."

With regard to the first objection, that the wires have to pass through the outer wall of high temperature, and that this will invalidate the reading, I think he is mistaken; there will be nearly an inch of heated wire between the thermo-couple and the point where the wires pass through the outer zone, and it seems to me that any conduction from these bodies will be more likely to take place outward along the cool wire than inwards to the heated zone, so that if an equal length of wire on each side of the weld is heated in order to prevent unequal resistances being introduced, the readings should not suffer much from this cause.

Nor does his second objection appear to me to be in any way a fatal one. If the gases in the inner zone had to pass through a circle or even a belt of high temperature, the objection would be valid, but, inasmuch as the gases are flowing up within a sheath of high temperature which exists even after the inner zone ceases to exist, and which is fairly constant in temperature, these gases will probably be as much exposed to radiation as the thermo-couple.

In taking such readings the thermo-couple is only exposed in the heated zone for a very short period, and it has been pointed out by Le Chatelier that the couple takes up the temperature of the locality in which it is placed with astonishing rapidity, whilst Mr. Callendar has pointed out (Trans. Roy. Soc., 1892, 166) that a spiral of plati num wire is a bad radiator, and is exceedingly sensitive to slight changes in temperature; it may be assumed, therefore, that it will not absorb radiant heat with great rapidity, and that the temperatures recorded in the non-luminous zone are not far removed from the truth.

Professor Smithells further accentuates his distrust of these recorded temperatures by saying "the readings are not even comparable, as the flame varies in breadth and thickness from point to point in a vertical plane."

The inner zone undoubtedly becomes narrower in the higher