LONDON. Royal Society, February 8.—Sir Archibald Geikie, K.C.B., president, followed by Sir Alfred Kempe, vice-president and treasurer, in the chair.—Sir Norman Lockyer: The spectrum of comet Brooks (1911c). In this paper an account is given of the lines shown in a series of ten photographs of the spectrum of comet Brooks, taken between September 6 and October 31. Seven of the photographs were taken while the comet was an evening object, and three when it was a morning object. The instrument used was a 2-in. quartz-calcite prismatic camera. In the best spectrum (September 30), in addition to the well-established carbon or carbon-compound bands at?? 3883, 4737, 5165, 5635, other radiations were seen at?? 3io, 316, 337, 405, 421, and 436. Line? 421 is probably the cyanogen band, the head of which is? 4216. So far as is known, the ultra-violet bands?? 310, 316, 337 have not been recorded in the spectrum of any previous comet. Attempts have been made to ascertain the chemical origin of these lines by reference to published records of laboratory spectra, and to recent photographs of the spectrum of CO taken with the quartz-calcite prism, but with no success. Although no definite changes in the relative intensity of the cometary lines were noted amongst the earlier photographs, a comparison of the best of these (September 30) with that of October 31, when the comet was a morning object, showed the following changes:-(1) On September 30 line? 4216 was weakest of the three subsidiary lines?? 405, 4216, 436. On October 31 it was strongest. (2) Lines?? 3883, 4737 were of about equal intensity on September 30. On October 31? 3883 was distinctly the stronger. (3) The ultra-violet lines?? 310, 3l6> 337, shown in the spectrum of September 30, were not seen on October 31. A photographic comparison is given of the Kensington spectrum of comet Brooks (September 30) with that of comet Daniel (191 id), reproduced by Campbell in Lick Bulletin No. 135. Although the latter showed farv more detail, being photographed with a slit spectrograph, it is fairly evident that the spectra of the two comets are very similar.—Hon. R. J. Strutt: A chemically active modification of nitrogen, produced by I the electric discharge.—III. (1) Active nitrogen emits its I energy more quickly, and reverts sooner to ordinary nitrogen, if it is cooled. This is apparently a uniqueinstance of a chemical change accelerated by cooling. (2) If the glowing gas is compressed to small volume, it flashes out with great brilliance, and exhausts itself in so doing. This proves that the glow-transformation is poly-molecular, i.e. that more than one molecule must take part in it. (3) Active nitrogen may revert to ordinary nitrogen in two distinct ways. One of those is a volume change, accompanied by glow; the other a surface action of the walls of the vessel, without glow. This is analogous to the behaviour of oxyhydrogen gas in its transformation to water, which may be a surface or volume effect, according to circumstances.—R. Whytlaw-Gray and Sir W. Ramsay: The atomic weight of radium. The material for this research consisted of 330 mg. of a mixture of radium and barium bromides, containing 206 mg. of radium bromide, supplied by the courtesy of the British Radium Corporation. The bromides were submitted to methodical fractional crystallisation, and yielded specimens of which the change in weight on conversion from bromide to chloride with gaseous hydrogen chloride, and from chloride to bromide with gaseous hydrogen bromide, was determined with the micro-balance. The atomic weight increased progressively from 2207, through a series of approximations, to the final atomic weight 226-36, the last five determinations giving the figures 226-40, 226-25, 22635, 226-35, and 226-45. The paper contains remarks on the differences in terms of multiples of the atomic weight of helium between the recorded determinations of the atomic weight of uranium and radium on the one hand, and of radium and lead on the other, and it is pointed out that a careful revision of the atomic weights of lead and of uranium, especially of the latter, is much to be desired.—Dr. J. A. Harker and Dr. G. W. C. Kaye: The emission of electricity from carbon at high temperatures. This paper discusses several new phenomena, among which are the generation of electric currents of considerable magnitude by what appears to be a new method. Two insulated carbon electrodes are inserted into a carbon tube resistance furnace at high temperatures, and are connected externally through a suitable current-measurer. If one of the electrodes is suddenly displaced to a colder or hotter part of the furnace, a reversible transient current is produced in the circuit without the application of any external potential. By such means, currents up to 2 amperes have been obtained. The production of an alternating current is thus rendered possible by the use of a suitable periodic device. A continuous current can be generated by suitably modifying the apparatus so as to maintain a large permanent temperature-difference between the electrodes. A steady current of?8 ampere has been thus obtained for a few minutes, and o-i ampere for more than an hour by water-cooling one electrode. These currents are such as would be produced by a discharge of negative particles from the hot electrode. At the lower temperatures positive currents have also been detected, but of much smaller magnitude. All the observations were made at atmospheric pressure. The extent of the ionisation of the furnace atmosphere at high temperatures was such that quite small E.M.F.'s, applied to two exploring electrodes, gave rise to steady currents of relatively enormous magnitude. For example, with 8 volts, currents up to 10 amperes have been obtained at a temperature of about 25000 C. Some of these observations have been repeated with furnaces of a non-electric character.— Prof.?.?. Barnes: The so-called thermoid effect and the question of superheating of a platinum-silver resistance used in continuous-flow calorimetry.—Prof. E. G. Coker: An optical determination of the variation of stress in a thin rectangular plate subjected to shear. The distribution of stress in a rectangular plate subjected to shear is examined by observing the optical effects in polarised light produced in a stressed plate of xylonite. Measurements of the shear stress at a point are obtained by using a specimen of material similar to the plate, and set along the direction of principal compression stress. A tension load is applied to this member of sufficient amount to produce a dark field at the point under examination. The intensity of tension stress so produced is twice the density of the shear stress in the plate. A survey of the central longitudinal section of a long rectangular plate shows that the shear stress rises very rapidly from a zero value at the ends, and reaches a maximum at a distance of rather less than the face width of the plate. As the distance from the ends increases, the stress decreases slightly in value until it reaches a minimum at the centre. A similar distribution also occurs at sections parallel to the central line. As the length of the plate is diminished the maximum and minimum stresses become more pronounced, and when the ratio of length to width is in the neighbourhood of two, the distribution changes in such a manner that there is a maximum at the centre. It is shown that the distribution is approximately parabolic when the length is equal to the width of the plate, and that when the length is greater than this the approximation is less close. The experiments show that a parabolic distribution of shear is only true within narrow limits, and that in a long rectangular section the distribution may be approximately represented by a uniform shear over the central section with a rapid fall towards the ends.—Dr. P. V. Bevan: Spectroscopic observations: lithium and caesium. -Captain C. F. U. Meek: A metrical analysis of chromosome complexes, showing correlation between evolutionary development and chromatin thread-widths throughout the animal kingdom. Measurements of chromosomes in organisms representing the principal phyla and classes of the animal kingdom have shown that lengths appear to constitute members of a series in arithmetical progression, whereas three distinct diameters exist, viz.?2?? in Protozoa, and 0-42? and 0-83? in low and higher meta-zoan phyla respectively. Consideration of these results has suggested the following working hypothesis:-The chromatin granules of simplest Protozoa are a visible expression of differentiation and aggregation of specialised particles concerned with transmission of hereditary characters, and as such probably do not represent the sole bearers of heredity in the cell. The granules become converted into rods by purely linear growth accompanying evolutionary development and greater somatic complexity, and, since the rate of growth is not the same in all chromosomes, rods of various lengths are evolved. A stage in phylogeny is later reached when a maximum rod length has been attained, such limit being imposed by spindle mechanism or other physical conditions; when this occurs, chromatin units conjugate in fours, and the normal thread-width is thus doubled. The newly formed chromosomes then segment into spheres of the same diameter, and these are prepared to enter a new course of linear growth accompanying further development. The same process is repeated when the length-limit has again been reached, and in this manner the greatest thread-width has evolved. The absence of correlation between chromosome dimensions and somatic characters is explicable on such an assumption, which postulates a series of cycles in the course of phylogeny. The heterotropic chromosome alone does not belong to the general series, and its great breadth may eventually be shown to be due to conjugation of normal rods; it is probably undergoing some process of development or disintegration, and may or may not be the determining factor in sex.