gray: linear expansivity 253 



One important object of the experimental study that developed 

 the simple plug just described was to secure data for improving 

 the design of electric furnaces, especially in the direction of better- 

 ing the distribution of temperature. The return-flow, oil-heated 

 tube was used in order to secure uniform lateral heating, and 

 thus to avoid confusing the observations by the effect of such 

 irregularities as would be almost sure to exist in an electrically 

 heated helix, no matter how carefully wound. 



Previous experimenters have tried to secure uniformity of 

 temperature in an electrically heated air column by using the 

 central portion of a sufficiently long tube and by crowding the 

 windings near the ends or other places where heat was lost most 

 rapidly. In this way Daniel Berthelot 2 secured uniformity within 

 2° over the central 22 cm. of a tube 85 cm. long heated to 917.° 

 Jaquerod and Perrot 3 in a well-insulated furnace without crowded 

 windings heated a gas thermometer bulb 8 cm. long with varia- 

 tions of less than 2° at 1066? Waidner and Burgess 4 constructed 

 a black-body in which compensation for heat losses thru the 

 ends was secured in great measure by the use of a second inde- 

 pendent heating coil surrounding the first and projecting 8 cm. 

 beyond each end. The winding of the secondary coil was very 

 close about the ends and very open about the center. At 1244°. 9 

 this arrangement gave such remarkably uniform temperature 

 that there was no variation of more than one or two tenths of a 

 degree over a length of 12 cm. However, the method of varying 

 the distribution of heat supply to compensate for lack of uni- 

 formity in the distribution of heat losses suffers from two serious 

 defects: (1) It is a method of cut and try. (2) The proper 

 arrangement to secure uniformity at one particular temperature 

 does not give uniformity at any other temperature, necessitating 

 a repetition of the cut and try process for every temperature 

 desired. This is strikingly illustrated by one of the compensated 

 black-bodies of Waidner and Burgess. While the temperature 



2 D. Berthelot, Ann. Phys. et Chim. 26: 119. 1902. 



3 A. Jaquerod and F. L. Perrot, Archives des Sciences Phys. et Nat. 20: 45 

 and 57. 1905. 



4 C. W. Waidner and G. K. Burgess, Bull. Bur. Standards 3: 165. 1907. 



