346 



NATURE 



•[November 26, 1914 



surrounding air do not affect the order of 

 magnitude of the results, but for the best 

 insulators the time of application of the voltage 

 has a serious influence on the values obtained. 

 Some of the most useful results are : — Fused quartz, 

 special paraffin, and ceresin, all over lo** ; micas 

 from 10" to 10"; sulphur, lo" ; bakelite, lo" to 

 10"; glass, 10" to 10"; unglazed porcelain, lo"; 

 paraffined wood, lo" to lo" ; marble, lo" to lo* ; 

 and slate, lo' ohms per centimetre cube. One of the 

 most noticeable features of the table given by the 

 author is the great difference between the insulating 

 properties of different specimens of materials going 

 by the same name. 



A VERY interesting example of an experimental steam 

 engine and accessories has just been supplied to the 

 Marine School, South Shields, by Messrs. W. Sisson 

 and Co., Ltd., of Gloucester. An illustrated descrip- 

 tion of this engine appears in Engineering for Novem- 

 ber 20. The engine is of the quadruple expansion 

 type, and has four cylinders operating on four cranks, 

 set at such angles as to give the best balancing results. 

 The arrangements permit of the engine being run as 

 a triple-expansion or compound engine, as may be 

 found expedient for experimental purposes. The high- 

 pressure cylinder is jacketed, and is provided with 

 Meyer's expansion gear, operating on piston valves. 

 The first intermediate cylinder is also steam-jacketed, 

 and has a piston valve of the maker's special labyrinth 

 type. The second intermediate and low-pressure 

 cylinders have single-pc/rted, flat-faced slide valves. 

 Each cylinder has Stephenson's link-motion reversing 

 gear. The engine drives either a dynamo or a Heenan 

 and Froude dynamometer. The air-pump is of the 

 ordinary single-acting type, and the surface condenser 

 is independent of the engine. 



A GREAT deal of experience has now been acquired 

 regarding the cost of operating and maintaining 

 systems of substations required in electric railway 

 installations, and Dr. H. F. Parshall read a paper at 

 the Institution of Civil Engineers on November 17 

 having for its object the assisting towards the 

 standardisation of electric railway substation practice. 

 The number of independent variables when a complete 

 system with substations has to be dealt with is so 

 great that the mathematical expression, from which 

 might be deduced the minimum cost, would in practice 

 be open to some suspicion. Hence the author gives 

 in the paper a complete balance-sheet embodying every 

 item for each case; without any great amount of 

 labour, the methods and results given may be applied 

 to practically any class of electric railway installation. 

 Curves are also given showing the arrangements of 

 substations that will operate different train services on 

 different electrical systems, and at various voltages 

 with a minimum total operating cost. Curves are in- 

 cluded illustrating the advantages gained by working 

 at high voltages, and these confirm Dr. Parshall's 

 view that with the present arrangement of rotary- 

 converter substations, there is little advantage in a 

 higher voltage than 2400 for the track conductor. 



Erratum. — On p. 318, col. 2, line 9 from bottom, for 

 "miles" read "metres." 



NO. 2352, VOL. 94] 



OUR ASTRONOMICAL COLUMN. 



Comet News. — Comet Delavan (1913/) is a morning 

 object but situated low down. The follov/ing two-day 

 ephemeris has been communicated by the Observatory 

 of Copenhagen, and is a continuation of that which 

 appeared in the Astr. Nachr., No. 4756 : — 



Ephemeris 12/1. Mean Time Berlin. 



Nov. 



Dec. 



Mag. 



Last week the elements and ephemeris of Lunt's 

 comet (1914^) were given in this column. It seems 

 that this comet was discovered independently by several 

 observers, and the Lick Observatory Bulletin No. 

 262 gives the elements and ephemeris of Comet 

 19146 (Campbell), which refers evidently to the same 

 object. In this bulletin it is stated that the first tele- 

 graphic announcement from the Harvard College 

 Observatory described it as a bright comet seen in 

 the southern constellation Doradus by Leon Campbell 

 at Arequipa. As the ephemeris given in this column 

 last week differs slightly from that printed in this 

 bulletin, a portion of it is reproduced below :— 



Greenwich Mean Midnight. 



In Knowledge for November, Dr. Crommelin gives 

 an interesting account of former appearances of 

 Encke's comet, together with a chart illustrating the 

 approach of this object to Mercury about twelve days 

 after its perihelion passage, which occurs on Decem- 

 ber 5 next. The comet, he states, is likely to be a 

 conspicuous object in November. The following is a 

 portion of the ephemeris which he publishes : — 



Ephemeris for Berlin Noon. 



R.A. Dec. 



h. m. s. , 



Nov. 26 .... 14 41 46 ... —10 16 



30 ... 14 57 33 ... 14 20 



Dec. 4 ... 15 17 20 ... 18 3 



8 ... 15 41 12 ... 21 18 



12 ... 16 7 27 ... 23 54 



Dec. 16 ... 16 34 6 ... —25 50 



Mr. Denning, in the Observatory for November, 

 communicates a short article on Encke's comet, and 

 gives a table of all the comet's returns, with the dates 

 of its redetection and periods. 



Spark Spectrum of Nickel under Pressure. — 

 Attention has often been directed in this column to 

 many investigations regarding the fact that the wave- 

 lengths of spectrum lines, determined from measure- 

 ments in the solar absorption spectrum, display small 

 differences when they are compared with the measure- 

 ments of the same lines in spectra secured from 

 terrestrial light sources. Numerous observers have 

 attributed these discrepancies to the effect of pressure, 

 but more recently Mr. Evershed has ascribed the solar 

 displacements to Doppler effects. It is of importance, 

 therefore, to pursue the study of the effect of pressure 

 on the behaviour of lines in spectra, and in this con- 

 nection the investigation of the spark spectrum of 

 nickel which has been completed by Mr. E. G. Bilham 



