April 20, 19 16] 



NATURE 



169 



on the subject and comes to the conclusion that the 

 most probable value of the constant is 575x10-'^ 

 watt, cm.-^ degree-*. 



The (ximposition of the exhaust from liquid-fuel 

 engines has been studied by Mr. R. W. Penning, who 

 presented a paper on the subject to the Institution of 

 Mechanical Engineers on March 17. Various fuels 

 were employed, the considj^rations affecting their 

 choice being volatility, purity, and general suitability 

 for use as motor spirit; hexane and benzene were 

 taken as standard, high-grade petrol and benzol as 

 commercial fuels. Mixtures with air were exploded 

 in a small glass vessel, and a complete chemical 

 analysis was made of the products. Exhaust samples 

 were also taken from an engine fitted with Dr. Wat- 

 son's apparatus for measuring air and fuel, and these 

 samples were analysed. In each case a set of curves 

 was plotted, taking as abscissae ratios of the fuel to 

 air by weight, and as ordinates percentages 

 of each of the products of combustion in 

 turn. Such a set of curves is termed an 

 exhaust-gas chart. The author concludes that with 

 volatile fuels there is little difference in the composi- 

 tion of the products of combustion from air-fuel mix- 

 tures in a small explosion vessel or in an engine 

 cylinder in spite of the conditions being so dissimilar. 

 I Another conclusion is that a very small quantity, if 

 ; any, of unsaturated or saturated hydrocarbons is pre- 

 1 sent in engine-exhaust gases ; this statement is, of 

 ' course, based upon the results obtained in gas analysis 

 by the method adopted and described. 



Ak important paper by Dr. C. H. Desch, on "The 

 I Decay of Metals," is included in a recent issue of the 

 I Transactions of the Institution of Engineers and Ship- 

 I builders in Scotland (vol. lix., part 5). Three chief 

 'types of decay are considered. The first is that due 

 I to allotropic change, of which the "tm pest," studied 

 I by Prof. Cohen, of Utrecht, is the most notable 

 example; similar disintegration may, however, occur 

 in certain light aluminium alloys, which are liable 

 to fall to powder as a result of internal molecular 

 change, though fortunately this does not occur with 

 any of the alloys in common use. Disintegration 

 jmay also occur as a result of internal strain set up 

 by hard working. Thus drawn rods are in a state 

 k)f severe tension in the outer layers, and in com- 

 pression in the inner layers, whilst in rolled or 

 [hammered rod this distribution of stresses is reversed. 

 The fracture of the strained metal may be accelerated 

 py corroding agents, which in some cases cause it to 

 "'-ack with almost explosive violence, as when very 

 : d-drawn rods of brass or bronze are touched with 

 ^ solution of a mercury salt or of ferric chloride, 

 finally, metals may decay as a result of actual corro- 

 sion, as in the " graphitisation " of iron pipes, from 

 vhich all the f errite is removed, leaving a soft residue of 

 :ementite, phosphide, and graphite. All these different 

 ypes of decay are illustrated by photographs, the 

 l-xcellence of which has become almost a commonplace 

 ture in the work of the author. 



, Mr. Wm. Shackleton, assistant inspector of 



.|cientific supplies at the India Store dep6t, writes to 



.irect attention to the numerals designed by his pre- 



essor, Col. A. Strange, F.R.S., in the early 



venties. These are still used on surveying instru- 



■ iients of to-day. Mr. A. P. Trotter is a nephew of 



'fol. Strange. He illustrated these numerals in the 



irnal of the Institution of Electrical Engineers for 



'^r^^/T I, gave details of their dimensions, and 



a them as a basis for his attempt to design an 



nroved set (see Nature, Februarv 24, p. 714, and 



nl 6, p. 121). - ^' ^ ' '*' 



NO. 2425, VOL. 97] 



OUR ASTRONOMICAL COLUMN. 



Comet 1916a (Neujmin). — The following elliptic 

 orbit has been derived by collaborators in the Berkeley 

 Astronomical Department (Lick Observatory Bulletin, 

 No. 280V from observations on February 29 (Yerkes), 

 March 8 (Bamberg), and March 7 (Lick) : — 



T = March ir2i95 G.M.T. P = 5'i86 years. /x = 684"i4" 

 <o=\gf 44-1'. ^ = 0-55465. 



^=327° 388'. (whence </) = 33- 41' 1 1 -8". 

 /== 10° 296'. log a = 047658). 



The resulting ephemeris diverges in R.A. from that 

 calculated at Copenhagen ; thus interpolation for April 

 20 gives o loh. 3m. 58s., and 6 — 7° 38-0'; the Copen- 

 hagen position being o loh. 2m. 41s., and 6 — 7° 34-0'. 



The comet was observed at the Hill Observatory, 

 Sidmouth, on April 8. It was then very diffuse and 

 faint, in approx. position at gh. 27- im. G.M.T., 

 o=9h. 39111. 51S., 8=— 3° 549', very nearly midway in 

 Right Ascension between the positions given by the 

 two ephemerides. 



Solar Radiation. — Mr. R. S. Whipple's paper on 

 instruments for the measurement of solar raciation, 

 read before the Optical Society of London on March 

 II, contains an account of all the most important 

 forms of instrument, from the Campbell sunshine 

 recorder and the black bulb in vacuo, to the register- 

 ing standard water flow pyrheliometer of the Smith- 

 sonian Institution. Of these instruments the Camp- 

 bell sunshine recorder still holds its own as one of 

 the most accurate means of measuring the duration 

 of sunshine, while the black bulb in vacuo, the read- 

 ings of which have been recorded so many million 

 times by patient observers, is now thoroughly dis- 

 credited. At the^ International Meteorological Con- 

 gress in 1905 the Angstrom pyrheliometer was adopted 

 as the standard instrument for the measurement of 

 the intensity of solar radiation. In this instrument 

 one of two similar metal strips is heated by the 

 radiation to be measured, the other by an electric 

 current seat through it. Equality of heating is 

 secured by two thermo-junctions behind the strips, the 

 necessary heating current is read, and the rate of supplv 

 of energy calculated. According to the most trustworthy 

 measurements made under conditions more favourable 

 to accurate observation than oiir climate permits, the 

 earth receives from the sun, on the average, 0032 

 calories per square centimetre per second. 



Proper Motion of the Orion Nebula. — M. J. 

 Comas Sol^ has obtained direct evidence that the 

 annual proper motion of the great nebula is about 

 0025" by stereoscopic comparison of photographs. 

 The near by small nebula, A.G.C. 1977, shows equal 

 motion, but in the opposite direction, and it is con- 

 sidered to be in orbital relation with the first. 



The System of A Tauri. — Prof. F. Schlesinger has 

 found that A Tauri most probably involves three main 

 bodies, only one being bright enough to yield a 

 spectrum (Publications, Allegheny Observatory, iii., 

 20). Partial eclipses at intervals of four days result 

 from the revolution of a less massive satellite, whence 

 also arises the chief oscillation of the spectrum lines, 

 but a second more remote and smaller body betrays its 

 existence and period of 346 days in a superposed 

 secondary oscillation revealed by the residuals. The 

 respective masses are largely conjectural ; on certain 

 assumptions they would be 25, 10, and 04 solar, and 

 the distances from the centroid of the first two 3-2, 8-0, 

 and 500 millions of kilometres. The great range of 

 velocity (56- 18 km.) found by Prof. Schlesinger, asso- 



