December 23, 1920] 



NATURE 



545 



the great local thickening of one of the coals, known 

 as the "Quarrelton thick coal," described as "cer- 

 tainlv the thickest single accumulation of coal-matter 

 in Britain," and the remarkable crush-zone known as 

 the ■' Paisley Ruck." Both these are fully and 

 exhaustively discussed in the memoir. 



SofTHPORT Meteorological Observations for 19 19 

 have recently been published by the County Borough, 

 the report being prepared by Mr. Joseph Baxendell. 

 Hourly whole-year normals of several elements for 

 periods ranging from ten to twenty years are given 

 in diagram form as a frontispiece. The tabular 

 statistics of air-pressure, temperature, sunshine, rain- 

 fall, wind direction and velocity, with other subsidiary 

 matters, are given for several hours for each month 

 and for the year with the differences from the 

 normal, kn outstanding feature of 1919 was the 

 remarkable prevalence of winds from the north- 

 westward, their duration surpassing that of winds 

 from any other direction, which is quite un- 

 precedented since the commencement of auto- 

 graphic records in 1898. The March of 1919 

 was the wettest experienced during the forty- 

 eight years' existence of the observatory, and this was 

 followed by a deficient rainfall for eight consecutive 

 mcnths. .Atmospheric pollution is regularly measured 

 at two different spots in Southport, and an attempt is 

 made to compare the results with those from other 

 places, but the observations to hand are by no means 

 numerous. The recorded meteorological observations 

 are of the highest order, and the report from vear 

 to year contains usually some special discussion— 

 a feature in advance of the returns from most cor- 

 (iorations. In the report for 1919 the special dis- 

 cussion is the diurnal variation of air-pressure from 

 the averages of ten years, and associated with this 

 i^ thi- r.ini^'c of the diurnal sea-breezes. Ten years 

 vv for such a discussion, and certain 

 in. iii> \A irregularity cannot be eliminated in so 

 ii>rt a period. 



The Risuini of the communications made to the 

 I'rench Physical Society on November 19 contains a 

 short description of the spectrograph for ultra-violet 

 ■Aork constructed by MM. J. Duclaux and P. Jeantet. 

 I he lenses of the collimator and camera are of quartz 

 and the prism is of water between quartz plates. The 

 whole is enclosed in a wooden box which is suflficiently 

 L;.is.tight to allow non-absorbing gases to be sub- 

 stituted for air. With a water prism of 70° it is 

 ssible to photograph spectral lines from 6000 to 

 sjo tenth-metres. For light of wave-length less than 

 2<>oo water is more transparent than quartz, and 

 throughout the whole of the ultra-violet the dispersion 

 of water is greater than that of quartz. With a con- 

 ' nser spark the line i86j may be photographed on a 

 - humann plate which is almost devoid of gelatine 

 5 seconds. The only inconvenience of the water- 

 ism for this work is the considerable change of th« 

 refractive index with temperature. 



Tub discovery of the nickeUsteefs of the invar 

 type, possessing a very small coefficient of expansion, 

 has had many valuable applications in metrology, 

 but careful study has proved the oxisleni-e of 

 N ' 2669, VOL. io6] 



residues, temporary or permanent, much resembling 

 the anomalies in the expansion of glass. The cause 

 of this instability has now been proved by M. Ch. Ed. 

 Guillaume (Comptes rendus, November 29) to be the 

 carbon in the alloy. The index of instability is 

 arbitrarily defined as the elongation, in microns per 

 metre, of a bar submitted to a heating for 100 hours 

 at 100° C-, the only previous treatment being the air- 

 cooling after forging. For an invar of average com- 

 position in the natural state this index has been 

 found to be of the order of 30. The examination of 

 a series of bars with gradually increasing and exactly 

 determined percentages of carbon showed that the 

 index of instability was a regular function of the 

 carbon content, and approached the origin of the co- 

 ordinates as the carbon approached zero. Cementite 

 was regarded as the probable cause, and hence the 

 addition of a metal having a higher affinity for carbon, 

 such as chromium, tungsten, or vanadium, suggested 

 a probable solution of the problem of removing, or 

 at least of reducing, the cementite. Experiments made 

 in this direction have proved successful, the addition 

 of chrwnium, for example, to an invar containing 

 o-i per cent, of carbon producing an alloy in which 

 the index of instability is reduced to one-tenth of its 

 original value. .As the price of the increase in 

 stability, the coefficient of exparl^ion is slightly raised. 



The meeting of the Illuminating Engineering Society 

 on December 14 was devoted to the presentation of 

 reports on progress and to exhibits illustrating novel- 

 ties in lighting. The number of committees working 

 under the society has been imreased, recent 

 additions being committees to d(?al with progress in 

 gas lamps and lighting appliances and with the 

 measurement of light and illumination. Progress in 

 the latter field was illustrated at the meeting by the 

 exhibition of a new form of illumination-photometer 

 enabling illumination to be determined by the 

 inspection of a series of apertures of graded bright- 

 ness without any manipulation being needed to secure 

 a reading. Such instruments seem likely to be specially 

 useful for demonstration purposes. Other featur.-s 

 of interest among the exhibits included a demonstra- 

 tion of the latest form of Sheringham "artificial diiy- 

 light " apparatus, which is now being made in a 

 standardised form, and some small floctrical gas-filled 

 lamps with opal glass bulbs, shown by the Kdison and 

 Swan Electric Co. This firm also loaned for exhibi- 

 tion some new neon lamps of a very interesting 

 character. Formerly such lamps, the light of which 

 depends upon an electrical discharge through rarefied 

 neon gas, required a relatively high voltage. But the 

 lamps exhibited resembled an ordinary glow lamp in 

 appearance, and could be insertinj in an ordinary lamp, 

 socket on 220 volts. The lamp exhibited was xtaled to 

 consume only 5 watts, but the light concists of a 

 relatively feeble orange glow at the negative electrode. 

 It is thought, however, (hat such lamps will be useful 

 as pilot lamps, etc., and in cas«'» where only a very 

 low candle-power is needed. At the conclusion of the 

 meeting the chairman (Dr. A. H. I^vy) nnnouncixl 

 that Mr. J. IferlK'ri Parsons had accepted the presi- 

 dency of the soiii ly in succession to Mr. \ V. 

 Trotter. 



