252 



SCIENCE 



[N. S. Vol. XXXIV. No. 



The physical properties of bonded alundum re- 

 fractories are as follows: 



High Melting Point. — Between 1,950 and 2,100 

 degrees Centigrade. 



Eigh Thermal Conductivity. — 2.1 times that of 

 vitrified firebrick and 1.6 times that of porcelain. 



Low Electrical Conductivity. — Even at elevated 

 temperatures it is a better insulator than porce- 

 lain. 



Low Thermal Expansion. — This is linear and 

 varies between .0000085 and .0000059. 



Sigh Mechanical Strength. — This can be varied 

 and is dependent to a certain extent on the other 

 properties desired. 



Porosity. — This can be varied between wide 

 limits from impervious bodies to those having a 

 high porosity which can be used for the filtration 

 of liquids of any gravity. 



Filtering crucibles of the gooch type can be 

 used for filtering precipitates quantitatively with- 

 out any previous preparation and can be used 

 repeatedly. 



Extraction thimbles are made for all kinds of 

 extraction work and can be cleaned by igniting 

 over a gas burner. 



This material is especially useful in small wire 

 wound resistance furnaces, as it prevents over- 

 heating and corrosion of the resister and has high 

 thermal conductivity. 



Crucibles for analytical work of a general na- 

 ture such as drying and burning filter papers, coal 

 analysis, etc., have been found more rapid and 

 longer lived than porcelain. Crucibles for melting 

 platinum and high melting alloys, where no slags 

 are present, have been found very efficient. 



Combustion boats of alundum are useful in the 

 determination of carbon in steels, as the iron oxide 

 formed does not combine with the alumina at the 

 temperature of combustion. For very high tem- 

 peratures it has been found advisable to use a 

 lining of specially prepared carbon-free alundum 

 between the boat and the sample. When used in 

 this way the boats last indefinitely, as many as 

 500 combustions having been made in the same 

 boat. 



The Determination of Vanadium in Vanadium and 



Chrome-vanadium Steels : J. E. ' Cain. (Bull. 



Bur. Standards, 7, No. 3, p. 377; J. Ind. and 



Eng. Chem., 3, July, 1911.) 



Various errors in the usual methods for deter- 

 mining vanadium in steel are pointed out and in 

 a few oases methods for correcting or eliminating 

 these are indicated. A new method based on pre- 

 cipitation of the vanadium by cadmium carbonate 



followed by electrolysis, reduction and titration, 

 is described. 



Determination of Dust in Blast Furnace Gas: 



L. A. TOUZALIN. 



Dust determinations can be made in any blast- 

 furnace gas at any stage of its passage from the 

 furnace to stoves, boilers or cleaners, if proper 

 means are used to insure the correct rate of with- 

 drawing the sample from the gas main. When 

 samples are withdrawn through a sampling pipe at 

 a velocity less than that in the main, high results 

 will be obtained. Conversely, when the sampling 

 velocity is too great low results are obtained. By 

 means of the proper apparatus, described in the 

 paper, very satisfactory efficiency tests may be 

 run on a system of gas cleaners. This leads to 

 the development of changes in construction which 

 often have a remarkable effect on such efficiency. 

 The apparatus described and the method of op- 

 erating the same are in constant use at the South 

 Works of the Illinois Steel Company. 

 The Examination of Fir Oil obtained by Steam 

 Distillation of Douglas Fir: Hbnet K. Benson 

 and Maec Darrin. 

 The Wood Distillation Industry of the Pacific 



Northwest: Henry K. Benson. 

 Satfish Oil as a Paint Material: Henbt K. Ben- 

 son and Wallace Eshleman. 



Note on the Analysis of Nitrous Oxide: Warren 



E. Smith and Edwin D. Leman. 



On attempting to analyze nitrous oxide as sup- 

 plied in cylinders in the liquid state, we found 

 ourselves confronted with certain difficulties. Con- 

 secutive samples of the gas as drawn off from the 

 cylinders will vary somewhat in composition, ap- 

 parently for the reason that the impurities (oxy- 

 gen and nitrogen) are in solution in the liquid 

 nitrous oxide, and sample of gas as drawn may or 

 may not have reached a state of equilibrium with 

 the liquid. This evidently will depend on the rate 

 at which the sample is drawn, the length of time 

 elapsing between drawing samples, and various 

 other factors. Duplicate results can easily be ob- 

 tained from a sample large enough for several 

 analyses, but there is no certainty as to what such 

 a sample represents. Again there is a regular 

 progressive change in the composition of the sam- 

 ples as drawn from the cylinders. The impurities 

 escape at a more rapid rate proportionally than 

 the nitrous oxide in which they are dissolved, and 

 the nitrogen escapes faster than the oxygen. Be- 

 low are three sets of figures obtained from the 

 same cylinder of material. 



