. The methods described are suitable for preparing the pure compounds in quantities of 1 kg or more. For each of the hydrocarbons and also for n-heptane and 2.2,4-trimethylpentane we have determined accurate values for the boiling point, melting point, densities at 15" and 20" or at 20" and 25", refractive indices and molecular refractions at the same temperatures for the wave lengths 6563. 5876, 4861, 4341 A. W e have also recorded accurate values for the surface tension at 20" and for the parachor. The critical solution temperatures with aniline have also been determined.

Values of the molecular refraction, calculated according to the formulae of L o r e n t z -L o r e n z or of G l a d s t o n e -D a l e are given for five different wave lenghts for a number of normal and branched alkanes and for a number of cyclanes. Additive and constitutional influences on molecular refraction and parachor are discussed. Exact values for the refraction equivalents of CH,, C and H are given. Dispersions, specific and molecular dispersions have been calculated; the influence of homology and isomerism on the dispersion magnitudes is discussed.T h e temperature coefficients of the specific gravities and of the refractive indices ( I = 5876) have been calculated. Section 1.In this paper we shall discuss the additive and constitutional influences on the molecular refractions (calculated in accordance with the formulae of L o r e n t z -L o r e n z and G l a d s t o n e -D a l e ) . the specific dispersions, the molecular dispersions and the molecular parachors for the alkanes and cyclanes described in our first paper 1).In our previous paper we have determined the refractive indices for the wave lengths 6563 (hydrogen a line), 5876 ( weights C = 12.00 and H = 1.008.

The proof given by K r a f f t for the terminal position of the double bond in cetene is not complete, as during the formation of cetyne-1 from cetene dibromide the possihility was not precluded of a rearrangement under the influence of powdered KOH.In the present paper it is demonstrated that: 1. This rearrangement does not take place: 2. During the preparation of cetyne-2 from cetene dibromide under the inhence of alcoholic KOH cetyne-1 is the first reaction product.This conclusively shows that cetene is hexadecenc-1 .By heating cetyl palmitate in vacuo K r a f f t 1) obtained an unsaturated hydrocarbon C1GHs2, which he called "cetene" and to which he attributed the structure of hexadecene-I. K r a f f t gives the following physical constants for his cetene: b.p. T o prove the terminal position of the (double bond in his cetene, K r a f f t 2 ) first prepared the dibromide, the properties of which were: m.p. 13.5", b.p. 225-227O at 15 mm pressure, and from this, by treatment with solid caustic potash (by which two molecules of hydrogen bromi*de were eliminated) he prepared cetyne-1 (b.p. 155O at 15 mm. m.p. 15'. dI5 = 0.7999), an acetylenic hydrocarbon with a terminal triple bond, since it produces a silver compound. The formation of this compound does not warrant the simple conclusion that the double bond in cetene also occupies a terminal position since, if treated with alcoholic potash, the same dibromide yielded cetyne-2 (with a nonterminal acetylene bond). K r a f f t also found that alcoholic potash converts cetyne-1 into cetyne-2, while prolonged heating with sodium converts cetyne-2 into cetyne-1. It remained to be shown, therefore, that solid potash does not bring abou conversion of cetyne-2 into cetyne-1.

A. From the determination of the freezing points of mhcxadecane, n-hexadecene-1 (cetene and 8 mixtures of these substance it appared that the freezing point curve in this system can be represented by a very weakly curved line, so that this condensed system consists of an uninterrupted series of mixed crystals. T h e crystals that are in equilibrium with the mother liquor are somewhat richer in hexadecane than the liquid phase.The addition of hexadecane to cetene raises the freezing point of the latter.

Dans plusieurs communications dans ce Recueil ') fut decrite l'etude photochimique de la reaction entre les cetones aromatiques et les alcools. La quantite de pinacone formee pendant les experiences en I'absence d'oxygene ou la quantite d'oxygene absorbee durant les autres experiences nous fournissait les moyens de mesurer les vitesses de reaction. Pour les details de cette mesure nous devons renvoyer aux publications citees et a quelques autres qui seront publiees bient6t.Apres avoir observe que l'aldehyde acetique peut Ctre oxydee par l'oxygene en forme de gaz (surtout dans la lumiere) et donne l'acide acetique, nous avons conclu que la determination de l'oxygene absorb6 pourrait Ctre contrdee par un dosage direct de l'aldehyde acetique presente a la fin de l'experience. Ce dosage devrait remplir les conditions suivantes :1. O n doit determiner avec une precision suffisante des quantites minuscules d'aldehyde.2. Les autres substances volatiles presentes dans le melange de reaction ne doivent pas troubler le resultat de la determination. Les reactifs et liquides titres doivent