The Band Smctrum of Ph y s i c a1 C h e m i s t r y. -Mercurv. ERNST LIESE (zeitsch. wiss. Photochem., 191% 11, 349--375).-Th; banded spectrum of mercury is readily obtained when the uncondensed discharge from a n induction coil is passed through a discharge tube in which mercury is distilled from one end to the other. I n addition to the bands described by Eder and Valenta, which are situated at the more refrangible end of the spectrum, the author has observed other bands stretching from the

more »

... een to the red end. These bands appear to be of an intermittent character, and although visual observations indicate that their structure is similar to that of the bands at the violet end of the spectrum, it has not been possible to obtain satisfactory photographs of them. Wave-length measurements have been made in the case of four of the more refrangible bands, and the band structure thus indicated is discussed in reference to the formuls of Deslandres and Ritz. I n neither case do the calculated frequencies agree satisfactorily with those corresponding with the series of lines actually observed in the banded spectrum. H. M. D. Magnetic Resolution of the spectre of Nickel, Cobalt, and Iron. (Mlle.) I. M. GRAFTDYK (Arch. N6erZand, 1912, [iiiA], 2, 192--220).-The resolution of the lines which occurs when a VOL. civ. ii. magnetic field is applied has been investigated in a quantitative manner. For iron, the spectral region examined extends from X = 4500 to A = 6500, f o r nickel from X = 2300 to X = 6300, and for cobalt from X=3300 to X=4200. If H denotes the strength of the applied field, then dX/h2H affords a measure of the magnetic resolution, and if this is plotted against the number of the resolved lines, a curve is obtained which represents the frequency with which a given effect is produced. The great majority of the lines give rise to triplets, but the quantitative data afford no evidence of the existence of definite series of lines. I n regard to the above-mentioned frequency curves, which are very irregular in character, there appear to be maxima for values of dh/X2N equal to 10.75 and 14-25 x in the case of iron, and when this quantity is equal to 10.50, 11.00, and 13-25 x lO--l3 in the case of nickel. Between these values and the normal resolution of 9 . 4~ 10-13 there appears to be no simple connexion. The frequency curve for cobaltl shows no very pronounced maxima. la. M. D. Norma1 and Anomalous Zeeman Effects. FRIEDRICH PASCHEN and E. BACK (Ann. P?iysik, 1912, [iv], 39, 897--932).--The resolution brought about by a magnetic field has been examined in the case of certain oxygen, helium, hydrogen, lithium, sodium, and potassium lines. The observations are discussed with reference to the general relationships which have been found in connexion with the magnetic resolution of series lines. H. M. D. A New Proof of the Existence of Molecules. V. NILS PIHLBLAD (Zeitsch. physikal. Chem., 1912, 81, 417-430. Compare A., 1909 1910, ii, 946).-A continuation of the work of Svedberg and Pihlblad. Colloidal solutions, suspensions, and molecular solutions of coloured dyes were made containing particles of varying sizes, and the light absorption investigated, The substances used were aniline blue-ZB, indophenol-l?, azobenzene, Bordeaux-R, and fast-red-A. The methods adopted for varying the size of the particles were : (1) addition of alkali, (2) addition of a protecting colloid (for example, rubber), (3) progressive mechanical sub-division, and (4) ageing the solutions, It is shown that here as in the former work (Zoc. cit.) that the light absorption curves of the disperse systems approach those for molecular solutions as the size of the particles is decreased. With increasing size of the particles the absorption curves show a flattening. Relationship between Fluorescence and Chemical Constitution. JOHAS NES STARK (Zeitsch. EZckt~ochern., I 91 2, 18, 101 l--lOl4).--Polemical, and contains a survey of the publications of H. Kauffmann made at the "Deut. Bunsen-Ges. Vers." held a t Karlsruhe (ihid., 481). The author opposes the view that fluorescence is a property depending in any way on molecular constitution, and cites evidence to show that it is rather an atomic property. The interpretation of the luminous effects obtained, by Kauff mann, in gases by the use of Tesla oscillations is adversely criticised.