70 MODERN SCIENCE READER 



to buy good alloys unless one is willing to study them 

 sufficiently to know how to treat them and then to supply 

 adequate facilities for so doing. 



It is not to be expected that small users will install com- 

 plete testing laboratories, but a few dollars invested in 

 having occasional tests made will be well spent. There 

 are, however, many large concerns that could and should 

 spend, say, $5,000, for which it is believed the whole or a 

 large part of the following equipment could be obtained: 

 The ordinary tensile machine, a microscope, electrical or 

 gas furnaces capable of fine regulation, a good pyrometer, 

 preferably recording. The tensile machine can also be used 

 for making Binnell hardness tests, spring deflection tests, 

 etc. In addition to these some form of drop testing ma- 

 chine, such as the Fremont, will be found valuable ; a vibra- 

 tory or repetitive impact test is nowadays considered a 

 necessity, while cold bending and torsion apparatus is use- 

 ful. This equipment will be of small use, unless a thor- 

 oughly good man is put in charge a careful, conscien- 

 tious man of sound judgment. This man should direct 

 the heating operations in the factory ; he should construct 

 furnaces which heat uniformly, and he should exercise 

 eternal vigilance in keeping the pyrometric installations 

 up to par. 



The best pyrometer of the thermo-couple type 1 should 

 be looked over and calibrated at stated intervals, espe- 

 cially if in constant use. Protecting tubes should be fre- 

 quently examined and renewed, and electrical contacts 

 looked over. Occasionally check up the millivoltmeter. 

 Unfortunately there are many pyrometers of the thermo- 

 couple type on the market which cannot be watched too 

 closely. Quite recently I visited two concerns where they 

 were hardening the same grade of steel and doing it well. 

 I a~:kod each concern what temperature it was using. One 

 said 1,300 F. and the other 1,700 F. The actual tem- 

 perature in both cases was probably 1,500 F. Another 

 1 Compare Fig. 2 in ' ' Why a Flame Emits Light. ' ' 



