T2% 



NA TUKL 



[November i;, 1923 



pass in and out of paralkl 



(3) 



\'m 



(liHicul 



added 



bftwcc 



luive Uan «il).-»fi viti to txjil" 

 and hnvf also liccn seen ti) I 



wii !> 1 1 1 i( .il 1 1 .iiiiu.i \ 



^ati ' .lil.ili.il !■ :: "I ! ■. 



.i.wi lie then |M 



ticN .jil.iii.it iim. t 



; in liiri, iiR-re is rcalK 

 II a brush discharge and th 

 iiirntK' ilcscrihed. Thr niiK 



.1 !m11 lujitnin- li.is 

 iiiil.ir Ir. litiiinj^ 



'' :. ' ,lt thr end 



than tlie 



,M .Mi. ollir nl the 



w hii h iithir-. I .111 tie 



iKiihiiiL' \ii\ similar 



If hall (il ,::hi\sing gas 



physical i)h< numciia 



yet produ' f d in a ' ' ^ * ry at all approaching hall 

 lightning: i^ tht iirogcn studied by Lord 



Ra\I I ahc we have a mass of nitrogen 



subji ' Irical disrhargc which continues to 



glow i«*i M>ua: iiiite after it has In-en removed from the 

 field. Lord Rayleigh, however, is unable to accept this 

 explanation of ball lightning, and all that we are able 

 to say is that adive nitrogen is the nearest physical 

 phenomenDii to ball lightning yet produced in our 

 laboratories. iJall lightning apjK'ars always to be 

 associated with a thunderstorm, and it is possible that 

 the intense discharge of a lightning flash can produce 

 some at I 111 I i' ( hange in the air or rain through which 

 the dis( harge passes. If this is so, the glowing matter 

 of ball lightning ma\ Ix in a state otherwi.se not met 

 with in Nature. 



Unusual Forms of Crystallisation of Cementite in Steel. 



r^EMENTI 



^■-^ iron tl 



11',, the larhiilr ot iron, whirh confers on 

 ic properties ol steel, exists in three 

 prim ipal lorni^ in li\ pereiileetoid steels, ( i ) the pseudo- 

 (lendritir hirni.(j) the cellular or interLianiihir form, 

 and [J,) the intrai;ranular Uinn which L;i\es ri-e to the 

 Widmann.slatlen struc ture. PseudoMkndritu distribu- 

 tion arises directly troni the irregular concentration of 

 the .solid solution which results on solidification. The 

 cellular varietv occurs between the szrains. i.e. in the 

 network of tlie L:rain junctions, while the W'idinann- 

 stiittcn structure is caused b\ the [irecipiialion of 

 cementite in the interior of the "grains thenisehes and 

 shows evidence of the directive intluence of the crystal- 

 line network of each grain. 



A. M. Portevin has examined a sample of steel 

 which has enabled him to make certain new obser\a- 

 tions in regard to these forms of cementite. These 

 results were presented at the autumn meeting of tlie 

 Iron and Steel Institute held recently in Ital\ . Thi 

 sample was found in the hearth of a blast-furnace, and 

 its exterior presented the characteristic concave facets 

 peculiar to intergranular fracture. The grains of which 

 it was composed were exceedingly well developed, their 

 size being of the order of i cm. in transverse thickness 

 and several centimetres in length. The specimen con- 

 tained 1-22 per cent, of carbon, 1-35 of silicon, and 0-17 

 of phosphorus. It was, therefore, very distinctly hyper- 

 eutectoid and corresponds, so far as carbon percentage 

 is concerned, to a fairly hard cutting tool. An ex- 

 amination of the microstructure of this sample revealed 

 the presence of the cellular and Widmannstatten modes 

 of distribution of cementite, but the pseudo-dendritic 

 form was absent. 



Intragranular Cementite. 

 A micrographic section usually shows the cementite 

 in needles arranged along three or four directions in 

 each grain. This corresponds spacially with lamellae 

 parallel with the faces of the octahedron, and has the 

 appearance which cementite assumes more particularly 

 in case-hardened samples very high in carbon. In the 

 sample examined by Portevin a different orientation of 

 the intragranular cementite was observed. The con- 

 stituent was present, not in the usual isolated rectilinear 

 needles, but in the form of bundles of numerous verv 

 small needles, or of groups of elements crowded to- 

 gether. These were apparently elongated prisms 



NO. 2820, VOL. I I 2] 



analogous to tlie iiri^nioid- ot Belaiew, grouped in 

 masses. This is apparently the first time that intra- 

 granular cementite has been noticed with these morpho- 

 logical characteristics. It can, however, also be pro- 

 duced in steel which has been strongly case-hardened 

 at a very high temperature and ver>- slowly cooled. 

 Inclusions and notably bubbles constituted < enlres of 

 crystallisation around which the bundles !■ - 



were grouped. 



InTERGKAM I.Ak ( I.Ml.NI n 1.. 



This is customarily described and represented as 

 en\-eloping the grains and appearing in a section as 



continuous ribbon-like filaments which do not displav 

 any < haracteristic shape or orientation. Howe and 

 Le\\ . howexer. ha\e directed attention to the needle 

 points which ini])inue ironi the lementite network into 

 the interior ot the urains. and have raised the question 

 as to whether these lake their direction in obedience 

 to the crystallisation orientation of the adjacent grain 

 or of that of the network itself. They have suggested 

 that both influences manifest themselves, and that 

 sometimes one and sometimes the other predomi- 

 nates. In the present sample there is no continuous 

 network of cementite surrounding the grains. There 

 is a grouping of this constituent along the confines of 

 the grain joints, the xariable orientation of which can 

 sometimes be attributed to that of the intragranular 

 elements of cementite dispersed within each grain and 

 sometimes appears distinctly different. In other words, 

 the two influences remarked In Howe and Le\ •. 

 manifest themselves. Fig. i represents the appearance 

 obtained after oil-quenchine at 050° C. followed by 

 annealing at 550 ( 'lich causes the 



great bulk of the p ntite, and mon 



especiallv the \\idmann.-taien i emeiilite, to disappear. 

 The photograph has been taken at the junction of 

 three grains. The needles which comjxise the network 

 have in one in.;t inee different directions in regard to 

 each Lzrain. e jiin. ti.ni the a]ipearance of the 



barbs ot a le.mui. while in tlie two other junctions 

 the\- ha\e an almost uniform orientation. It appears 

 that the structural elements of the network have dis- 

 tributed themselves along a mean direction or have 

 assumed a direction of their own. the influences of the 

 orientation of each urain conflicting with each other in 

 th'fe neighbourhood of the junction. The needles are 



