September i6, 1920] 



NATURE 



91 



should be replaced by one according to their shrink- 

 age and porosity. Most of the clays tested show 

 expansion at some part of the firing range, owing 

 probably to the allotropic changes in the silica 

 present. In general, clays with high alumina content 

 show a wide range of porosity, but there are notable 

 exceptions. From the tables of the properties of the 

 various clays given by the authors it is possible to 

 calculate the properties of a mixture of them when 

 fired to a given temperature. 



Continuing his researches on the alloys of iron 

 with chromium and tungsten. Prof. Honda has 

 recently published in the Science Reports of the 

 T6hoku Imperial University an investigation on the 

 structural constitution of high-speed steels containing 

 these elements. He concludes that in an annealed 

 steel containing 5 per cent, of chromium, 18 per cent, 

 of tungsten, and 06 per cent, of carbon the alloy con- 

 sists of a solution of iron tungstide in iron, together 

 with free tungstide and the carbides Cr.C and WC. 

 On heating such a steel above .Xc, the carbides dis- 

 solve, and the chromium carbide CrjC is converted 

 . into Cr,C, and metallic chromium. The higher the 

 temperature, the more the change proceeds in this 

 direction. On cooling, the reverse change takes place 

 only slightly, and the result is that at the ordinary 

 temperature a steel is produced containing the car- 

 bides, chromium, and the tungstide all in solid solu- 

 tion. This, according to him, is the constitution of the 

 hardened steel. The self-hardening property is con- 

 veniently studied by the temperatures of the trans- 

 formations, while the degree of tempering on later 

 heating is best studied by means of magnetic heating 

 curves. The tempering takes place in two steps — 

 one at about 4(x>° and the other above 700°. Prof. 

 Honda concludes that self-hardening and resistance to 

 tempering depend primarily on the quantity of Cr,C, 

 dissolved in iron containing chromium and tungstide. 

 These properties increase both with chromium and 

 carbon and with rise of temperature. The function 

 ■of tungsten appears to consist in lowering the tem- 

 perature, at which self-hardening begins to be mani- 

 fest. When this element exceeds 12 per cent, it exists 

 as fine globules of tungstide Fe,W, and these are 

 directly related to the cutting efficiency of the tool. 



We have just received from VV. Heffor and Sons, 

 Ltd., of Cambridge, a catalogue of their library of 

 second-hand books. Science and mathematics are 

 well represented by a number of the bigger text-books, 



; and several collections of bound volumes of scientific 



. journals are also offered for sale. 



I'liROfOH the omission of the word " hundred " 

 in line 28 of the first column of page 38 of last 



• k's Nature, the annual production of coal in 



• at Britain was erroneously stated to approach 

 three million" instead of "three hundred million" 



-MIS. It is shown in the article upon the proceedings 

 ' f the Section of Economics and Statistics of the 

 British .Association, printed elsewhere in this issue, 

 lh.it the output in ic\\y was 287,000,000 tons. 



-■'i;;, \-' ir.. 106] 



Our Astronomical Column. 



Tempel's Comet. — This comet is now fading, but 

 may still be visible for some weeks. The following 

 ephemeris, for Greenwich midnight, is by M. Ebell : 



Sept. 17 



Values of log r, log A: September 21, 0-2364, 9-9411; 

 October 15, 0-2737, 99730. 



.-\ photograph obtained at Bergedorf on .\ugust 15 

 showed a well-defined nucleus and a fan-shaped tail 

 which could be traced for about 1'. The corrections 

 indicated to Ebell's ephemeris were +235., S. 3-0'. 



Nova Cygni. — ^The position of this star referred to 

 the equinox of 1920-0 is R.A. igh. s6m. 24-775., N. 

 decl. 53° 24' 1-3'; annual precession, +1-50S., +9-7*. 

 Examination of past photographs shows no trace of a 

 star in this place on plates taken by Dr. Wolf and 

 Mr. Franklin .\dams some fifteen and twelve years 

 ago. They go down to mag. 17 and 15 respectively. 

 Two plates taken at Harvard on 1920 .August 9 with 

 a i-in. lens fail to show it, and it must have been 

 fainter than 9-5. .-V plate taken by Mr. Nils Tamm 

 in Sweden on .-Vugust 16 shows it of mag. 7, and a 

 Harvard one of .-Vugust 19 indicates mag. 4-8. Since 

 the maximum was not reached until .\ugust 24, the 

 rise in light occupied more than a week. The total 

 increase in light was at least 15 magnitudes, while 

 that of Nova .-Xquilae 1918 was only 11 magnitudes 

 (Harvard Bulletin, 729; Astr. Nachr., 5060). 



The Perth Section of the .\strographic Cata- 

 logue. — The publication of the great Astrographic 

 Catalogue has fallen far behind the expectations that 

 were formed when the scheme was initiated some 

 thirty years ago, but many new observatories have 

 stepped in to fill gaps left in the zones, and these are 

 showing much energy in pushing on their share of 

 the work. The Perth Observatory, under the direc- 

 tion of Mr. N. B. Curlewis, undertook the region 

 from -31° to -41°. The Catalogue will be com- 

 pleted in thirty-six volumes, each containing six hours 

 of R..\. in a single degree of declination. Vols. xvii. 

 to xxiv. have recently been issued. The following 

 table gives the number of stars in each voFume and 

 the ratio to the number in C.P.D. : 



PUi* 



cttitre 



36' 



■37* 



R.A. oh.-«h. 

 No. of V..- No. of 

 •un •'""' .ur. 



7.740 

 7,664 



47 

 44 



R.A. 6li.-iih. 

 Rati I 



2-5 



26 



R.A. I9li.-i8h. 



tt.r» •*•"" M». 



R.A. i8h.-i4h. 

 No. of 



2i.9»3 

 21,883 



Ratio. 



20,766 4-1 19,832 SO 

 20,667 35 M.562 38 



It will be seen that the ratio varies considerably, 

 being, on the whole, lowest where the star-density is' 

 greatest. 



The radius of the image is given for stars not 

 fainter than magnitude 8. The fainter stars have 

 their magnitudes indicated by a letter (from .\ to L) 

 referring to a speciallv constructed scale ; approxi- 

 mately \ is of magnitude 8J, and the letters are 

 half-magnitudes apart, so that I. is mJ. 



The ^toiles dc rcpcrc, of which tm-re are about 

 fifteen on each plate, have recently been re-observed 

 with the Perth meridian-circle, and proper motions 

 deduced where necessary. The measured rectangular 

 co-ordinates of all the stars are given to o-ooi of a 

 r^seau interval. The usual plate-constants and tables 

 for reduction to R..\. and declination are also given, 

 with a note that the constants are deduced on a some- 

 what difforrnt nLin from that followed at Greenwich. 



