September 24, 19 14] 



NATURE 



97 



relationships or periodicities. If an attempt is made 

 to correlate a pair of variable quantities there is a 

 certain measure of correlation to be expected even in 

 the absence of any definite relationship. This 

 measure must be sensibly exceeded if the result 

 actually found is to be treated as an indication of 

 causal connection. When, however, a whole set of 

 correlations is determined the fortuitous element will 

 occasionally be largely exceeded without indicating a 

 real relation. Similarly, when a complete periodo- 

 gram is investigated the largest ordinate must exceed 

 the probable measure by a large factor if the corre- 

 sponding periodicity is to be regarded with any con- 

 fidence. The outcome of Dr. Walker's paper is a 

 table giving the ratios of the greatest of an assigned 

 number of (i) accidental correlation coefficients, (2) 

 amplitudes of periodicities, to the probable value of a 

 single one. The note illustrates the need for extreme 

 caution in interpreting the results of work of this 

 kind. But it does not touch the real difficulty of 



•riodogram analysis. This consists, in the language 

 of the optical analogy, in the occurrence of banded 

 spectra. Until this feature becomes amenable to 

 sound interpretation, the periodogram method, always 

 painfully laborious, is likely to remain absolutely 

 sterile. 



From the recently issued report of the Government 

 Chemist (Cd. 7562) we learn that a large number of 

 determinations of the salinity of sea-water were made 

 last year at the Government Laborator}-, for the in- 

 formation of the Permanent International Council 

 dealing with the exploration of the sea^ More than 

 four thousand samples of sea-water were taken by 

 steamships on various routes and by lightships round 



he coast, and forwarded for examination. The data 

 relating to the specimens from the Atlantic routes are 

 published with the monthly charts issued by the 

 Meteorological Office. It is noted in the report that 

 the chemical work of the Geological Survey is in 

 future to be performed by the staff of the Government 

 Laboratory, and arrangements have been made for 

 this purpose. During the year an extensive series 

 of experiments was conducted in connection with the 

 measures to be taken for the preservation of the roof 

 of Westminster Hall ; the results are not stated, but 

 a summary of the conclusions arrived at has been 

 published as an appendix to the report of the architect 

 in charge of the work. There is an interesting refer- 

 ence by the Government Chemist to the composition of 

 a series of medieval wax seals, of various dates from 

 the thirteenth to the sixteenth centuries, which were 

 examined for the Record Office. An impression of 

 the Great Seal of 1350 was found to consist of pure 

 beeswax which, although nearly six centuries old, 

 corresponded exactly in properties with wax of recent 

 origin. Among other matters of interest mentioned 

 are some experiments upon the keeping of milk in 

 galvanised-iron vessels : the results showed that at the 

 end of twenty-four hours the milk had dissolved zinc 

 equal to about half a grain per pint, and also traces 

 of antimony and arsenic. There was a large increase 

 in the total number of samples examined last year as 

 compared with the previous year, the respective num- 

 bers being 234,7:;4 and 209,502. 

 NO. 2343, VOL. 94] 



Part v. of vol. xxvi. of the Proceedings of the 

 Physical Society of London consists of 114 pages of 

 papers communicated to the society, the index to the 

 volume and the proceedings at meetings of the society 

 during the session 1913-14. Readers of the Proceedings 

 will, we feel sure, cordially welcome the issue of them 

 with their edges cut. It will be noticed that the 

 society has a committee on nomenclature and symbols 

 which makes a provisional report on electric and 

 magnetic symbols in the present number. Communi- 

 cations dealing with electrical subjects show a marked 

 preponderance over those relating to other branches of 

 physics. Of the electrical papers may be mentioned 

 Mr. F. E. Smith's on a variation magnetograph, Mr. 

 F. Mercer's on the arc electric wave generator. Prof. 

 Fleming's on the bending of such waves round the 

 earth, and Sir Joseph Thomson's account of his search 

 for waves still shorter than those hitherto produced. 

 A paper by Mr. T. Barrett deals with the thermal and 

 electric conductivities of rare metals, and one by 

 Prof. Lees with Fourier series and the method of 

 least squares. 



Prof. F. C. Lea, of the University of Birmingham, 

 and Mr. O. H. Crowther give an account in Engineer- 

 ing for September 18 of some experiments on rein- 

 forced concrete beams. In these experiments the 

 longitudinal strains of a part of the beam subjected 

 to uniform bending moment were measured by a 

 special type of extensometer. From the resulting 

 cur\-es, it appears that the ordinary assumption that 

 the distribution of stress is linear on the compression 

 side of the neutral axis is approximately correct. The 

 experiments show also that the modular ratio 

 diminishes as the stress increases, and is slightly 

 higher for the richest mixture than for the poorer 

 mixtures. The evidence so far is opposed to the 

 recent London Count}- Council Regulations, but the 

 authors consider that it is desirable that the experi- 

 ments should be continued, and have made arrange- 

 ments for this to be done. 



i OUR ASTRONOMICAL COLUMN. 



j A New Comet. — It is announced that a bright 



I comet was discovered at the Cape Obser\'atorA' on 



; September 18. The comet was then near the star 



j Achernar (a Eridani), and therefore invisible in the 



I British Isles. 



j Comet 1913/ (Delavan). — Comet Delavan is now 

 I beginning to diminish more rapidly its northern de- 

 ! clination, but is still a circumpolar object, and there- 

 I fore visible throughout the whole night. It is a con- 

 } spicuous object to the naked eye, and lies nparly on 

 ; the prolongation of a line towards the horizon joining 

 ; the two pointers (a and /3) in Ursa Major. This posi- 

 I tion can be easily found by consulting the chart given 

 in this column for September 3. The comet is steadily 

 increasing in brightness, and its tail has considerably 

 increased in length. The dark nights recently experi- 

 enced have been verv favourable, both for visual and 

 photographic work, and no doubt a large harvest of 

 data is being collected. 



M. Bigourdan, in the Cotnptes rendus for September 

 7 (tome clix.. No. 10), gives some" obsenations for 

 position of this comet which he made at the Paris 

 Observatory- on September 2 and 3. 



