During two campaigns of smelting Pao-Tou iron ore in a 1 M~3 experimental blast-furnace,samples of furnace gas were taken at different levels and analyzed for fluorine and silicon. Inorder that absorption of fluorine from the furnace gas by the dust contained there-in shall not takeplace during cooling of the gas samples, a method has been devised for separating the dust fromthe furnace gas at the prevailing temperature. The dust samples were analyzed by chemical and X-raymethods. Besides, some solid samples were also taken from the furnace and analyzed chemically. From the results of this investigation, it has been found that the fluorine content of the furnacegas depends primarily on the temperature. Consequently, the fluorine content of the blast-furnacetop gas is expected to be extremly low when the top gas temperature is not unusually high. Infact, the top gas samples analyzed as low as 3—13 mg. of fluorine per M~3 under normal conditions.It has also been found that fluorine exists in the furnace gas predominantly in the form of HF.From a thermodynamic point of view, the reactions between HF in the furnace gas and CaO orCaCO_3 in the solid charge have been found to approach equilibrium closely. This fact explains whythe fluorine content of the furnace gas is controlled by temperature as a principal factor. From the distribution of fluorine among the various furnace products, it has been found that apredominant part of fluorine enters the slag, a small amount goes into the dust, while the fluorinecontent of the top gas is very low. From the evidence provided by X-ray investigation, it has beenestablished that fluorine exists in the dust chiefly as CaF_2. The mechanism of the movement of fluorine in the blast-furnace is briefly discussed.

An extraordinarily large gas pocket formed during the rolling of a rimming steel slab wasstudied. Results of investigation indicated that the serious lamination defects that often occurredin such slabs were due to the presence in large rimming steel ingots of central cavities that (1) con-tained gases at high pressure, (2) were surrounded by heavily segregated metal, and (3) wereloaded at their surfaces with non-metallic substances. During the rolling operation these cavitiesfailed to weld up. On the contrary, the solid walls dividing neighbouring cavities were pushedopen and, as a result, a large discontinuity remained in the rolled slab. It was found that aftershearing the slab and thus exposing the discontinuity to atmosphere, in most cases such discon-tinuities could weld up during subsequent rolling into plates. Lower tapping temperature and lower pouring temperature (speed) were found effective tolessen lamination rejects.

~~

The rolling and recrystallization textures in a 50 percent Ni-Fe alloy have been studied,the X-ray pole figure method being employed. Meanwhile, the formation of the cubic texture in thisalloy has also been investigated. The samples used in the experiments have been given two kinds of cold reduction; i. e., (Ⅰ) 94percent and (Ⅱ) 98.4 percent. The samples of class Ⅰ and class Ⅱ both have been annealed for onehour at 500°, 600°, 700°, 800°, 900°, 1000° and 1100℃. From the experiments, the formation of the cubic texture was found to begin at 600℃, be-coming intensified with the raising of the temperature. The cubic texture in the samples belongingto class Ⅱ, which has been annealed for one hour at 1100℃, was found to be sharp and complete. From the experiments it may be shown that the cubic texture in the 50 percent nickel-iron alloyhas been formed from the rotation of～(123) [121] orientation about the [111]-axes for 40°. Theresults seem to verify that the development of the recrystallization texture in the nickel-iron alloyconforms with the theory of orientation growth.

The D. C. and A. C. methods are used for the measurement of the magnetic properties of 50pct. nickel and 50 pct. iron Ni-Fe alloy. The ratio between the remanence and the magneticsaturation induction, B_r/B_s, and the coercive force, H_c, are taken as a criterion for assessing themagnetic characteristics of the alloy. The samples used in the experiments are cold-rolled with (Ⅰ) 94 per cent, (Ⅱ) 98 per cent and(Ⅲ) 99 per cent reductions. The class Ⅰ and class Ⅱ samples are annealed at (1) 900°, (2) 1000°, (3)1100°, (4) 1200° and (5) 1300℃ for one hour. The class Ⅲ samples are annealed at 1100℃ for(1) 30 minutes, (2) one hour, and (3) one and half an hour respectively. The formation of the cubic texture of the alloy has been studied by the X-ray pole figuremethod and the results Will be presented in another paper. The development of the material used for the magnetic amplifiers is finally discussed.

Relationship between carbon content and height of the internal friction peak of carbon diffusionin pure nickel have been measured. In addition to this, the relationship between height of theinternal friction peak and the carbon content in the manganese steel containing 18.5% manganesewas also accurately determined, the obtained curve shows that the intersection of axis indicating thesolubility of carbon for this steel really exist. So that, the internal friction peak induced by the interstitial atoms diffusion in the pure metalof face center cubic crystal was owing to the rotation of the atomic pair of two neighbouring atomsin the crystal lattice under the action of stress. The state of carbon atoms may be modified by ad-dition of alloying elements, it can considerably reduce the possibility of formation of carbon atompair, consequently the internal friction peak in the high manganese alloy steel is primarily due tothe rotation of carbon-manganese atom pair.

Determination of boron in steel and iron by the aluminum spark method possesses highersensitivity and precision. It is not necessary to separate iron from the sample. The sample isdissolved with perchloric acid, and into the perchlorate solution is added a certain amount ofNaOH solution. The standards are prepared from synthetic solutions with different proportionsof boric acid added. Electrodes are made from pure aluminum 8 mm in diameter and 5-10 cm in length. Theends of the electrodes should be machined flat and the sharp edges should be removed by light cham-fering to reduce the tendency of sparks to strike these sharp edges. Drops of solutions are placed onthe surface of the electrode by means of a micropipet calibrated to deliver 0.05 ml. With a con-trol spark source and medium spectrograph we can analyze boron in the range from 0.005％ to1.0％ by the line B 2496.778A.

Adding small amount of rare earth elements in alloy steel will improve its certain mechanicalproperties. Direct spectrographic determination of residual rare earth in steel has been generallyregarded as impossible since the complex spectra of the major components of ten prevented the identi-fication and measurement of the weak rare earth lines. In this work the chemico-spectrochemical method is employed. The mercury cathode separ-ation and the ether extraction technique were used for the concentration of R. E. proceeding thespectrographic procedure. Yttrium or thorium is used as carrier and internal standard. The concentrated sample solution is droped on the tip of spectra pure carbon electrodes. A.C. arc (-1)is used to excite the spectra. This method permits the determination of as little as 50 γ of total rare earths in 1 gram sampleof steel with an error of±10％.

In the spectrochemical analysis of high chromium steels, the calibration curve of chromium wasfound to be affected by the difference in carbon content and the microstructure of the samples.This effect became considerable only when carbon was present in the solid solution and its contenthigher than 1％. Experiments show that the presence of carbon affected the volatilization of the different sub-stances from the solid electrodes, and this effect on chromium was greater than on iron. Thus, webelieve in the solid solution the probability of the formation of a Fe-C-Cr bond was greater thanthat of the Fe-C-Fe bond.