662 



NATURE 



[February ii, 191 5 



associated with line spectra, and can be represented 

 by the usual formulae involving the Rydberg constant. 

 Nine bands of the main series, and four of the second 

 fainter series, have been identified. (2) The two 

 series may be likened to the principal and diffuse 

 series in the case of line spectra, but the usual rela- 

 tion between such series is not certainly indicated, 

 and no equivalent of the sharp series has yet been 

 traced. (3) The doublet separations are not in accord- 

 ance with those associated with line spectra ; they 

 diminish in passing along the series, but do not 

 vanish at the limit. No regularity in the arrange- 

 ment of the single bands has been recognised. — 

 T. R. Merton : The spectra of ordinary lead and lead 

 of radio-active origin. Wave-length measurements 

 have been made of some of the principal spectrum 

 lines of ordinary lead and the lead in pitchblende resi- 

 dues. It might reasonably be expected that small 

 differences in the wave-lengths would be found, more 

 ^specially in the case of any lines which belong to 

 doublet or triplet series, since, according to the views 

 of Prof. Hicks, an atomic weight term enters exactly 

 into the separations of such doublets or triplets. No 

 series have yet been discovered in the spectrum of 

 lead, but it is probable that they exist, and an esti- 

 mate of the order of the differences of wave-length 

 to be exj>ected, according to Prof. Hicks 's views, is 

 given. No differences of this order have been 

 observed, the spectra taken in juxtaposition showing 

 no differences of wave-length, whilst the wave-length 

 measurements of seven of the most prominent lines 

 in the two leads agree to 003 A.U. A special com- 

 parison of the line A = 4058 A.U, with the interfero- 

 meter shows that the wave-lengths of this line in the 

 ordinary and radio-active lead are identical to within 

 0003 .\.U.— A. O. Rankine : The viscosity of the 

 vapour of iodine. The paper records the method used 

 by the author for measuring the viscosity of iodine 

 vapour at four different temperatures ranging from 

 124° C. to 247° C. The basis of the method is the 

 distillation of iodine from one vessel to another 

 through a narrow capillary tube, the temperature of 

 which could be varied. One of the vessels contains 

 solid iodine, and is raised to 100° C, thereby estab- 

 lishing the driving pressure. The other vessel is 

 immersed in a freezing mixture, ' so that the iodirte 

 which is transpired as superheated vapour is con- 

 densed there, the mass so transpired being determined 

 by weighing the vessel, which is removable. The 

 rnethod gives, in repeated experiments, verv con- 

 sistent results, and the values obtained mav be' taken 

 as accurate to about \ per cent. The range of tem- 

 perature variation is not large enough to make the 

 results suitable for testing the validity of Sutherland's 

 formula, but it is found that they are, at any rate, 

 not inconsistent with that formula. 



Physical Society, January 22.— Sir J. T- Thomson, 

 president, in the chair.— Dr. A. Russell : Practical 

 harmonic analysis. Making the assumption that the 

 graph of a periodic function is given, the problem of 

 the best way of determining the Fourier constants in 

 the series equation which represents it is considered. 

 The ordinary method of procedure is to neglect all the 

 harmonics above a certain order and determine the 

 coefficients of the harmonic terms by making the 

 curve represented by this equation pass through a 

 number of arbitrarily selected points on the given 

 curve. This is the method used, for instance, by 

 Runge and Grover. A serious defect in this solution 

 is that the values found for .the amplitudes of the 

 harmonics, more especially for the higher harmonics, 

 may be very different from their true Fourier values. 

 The method gives no indicat;on of the magnitude of 



NO, 2363, VOL, 94] 



these errors. Gauss pointed out many years ago thai 

 the solution of this limited problem could be written 

 down at once mathematically, and that it was ol 

 importance in certain interpolation problems in astro- 

 nomy. Another method has been suggested recently 

 by Silvanus Thompson. He uses certain series 

 formulae for finding the Fourier constants. The 

 author suggests other series formulae of a similar 

 kind, and gives numerical examples to illustrate the 

 accuracy attainable by the use of infinite series 

 formulae. T. Smith : Measuring the focal length of w 

 photographic lens. The focal length of a compound 

 lens is obtained solely by focussing on the camerii 

 screen the image of a distant object on the lens axis 

 by the complete lens and by each of its components 

 separately. One- additional focussing of the same 

 object when the separation of the components is 

 altered determines the focal lengths of each com- 

 ponent. The method is both accurate and quick, and 

 requires only a camera and the lens. 



Challenger Society, January 27. — Dr. G. H. Fowler 

 in the chair. — D. J. Matthews: Hydrographical ob- 

 servations on the ice patrol Scotia in the North Atlan- 

 tic, 1913. The Scotia cruised in the ice-area from 

 April 9 to August 3, 1913, and crossed the Labrador 

 current fourteen times between the latitude of Hamil- 

 ton Inlet in 54° N., and the southern part of the New- 

 foundland Banks in 44° N. Observations of salinity 

 and temperature were made down to a depth of 550 

 fathoms, including seven complete sections across the 

 current. The Labrador current had a temperature 

 generally below 0°, except at the surface, and a salinity 

 of 32-5 to 335 per thousand, while the Atlantic water 

 to the eastwards reached 34-5 per thousand or more. 

 The boundary between the two water-masses was 

 marked by the isohaline of 3400 in the region north- 

 ward of the Banks; further south the conditions were 

 more confused. The characteristic temperature mini- 

 mum was found in all water of polar origin except on 

 the Banks, where it was disturbed by the shallows. 

 Measurements of currents and of the drift of bergs 

 were made by means of mark buoys, either anchored 

 to a sinker on the bottom or to a drag working at a 

 depth of 1000 fathoms. The velocities observed were 

 all low, generally less than half a mile an hour, and 

 calculations by Bjerknes's method from differences of 

 density gave similar results. Oft' Cape Race a slow 

 northerly set was observed instead of one of about one 

 mile per hour to the south and west, as is usually 

 experienced here. On the Banks measurements were 

 made with the Ekman current-meter at depths of five 

 and twenty-five fathoms at two stations ; the observa- 

 tions were repeated at frequent intervals dujring 

 thirteen hours, and showed that the current is almost 

 entirely tidal with a very small easterly resultant. 

 Prof, Barnes's observation that the proximity of an 

 iceberg can be detected by a slight rise in the tem- 

 perature of the water could not be confirmed, as similar 

 rises were recorded when no ice was near. — C, Tate 

 Regan : Colour-changes in a flat-fish, Platophrys 

 podas. 



Linnean Society, February 4.— Prof. E, B. Poulton, 

 president, in the chair. — Ruth C. Bamber : Report on 

 fishes collected by Dr. Crossland from the Sudanese 

 coast of the Red Sea. This collection consists of 

 ninety-one species, of which two are new to science 

 and two others require re-description. Most of the 

 others are typical Red Sea coastal fishes, and were 

 obtained by Dr, Crossland at various localities on the 

 Sudanese coast between Suez and Suakim. — Dr, 

 Marie C. Stopes : A fossil of doubtful affinity. The 

 fossil was found in situ in the Lower Greensand at 

 Luccomb Chine, in a horizon in which a number of 

 new plants as well as the famous Bennettites Gib- 



