108 PHENOMENA, ATOMS, AND MOLECULES 



so that the method was not applicable for iio-volt lamps with less than 

 500 watts. Some time later, however, it occurred to me that the benefits 

 derived from the large diameter of the filament could be obtained with 

 one of smaller diameter by coiling the filament in the form of a helix, bring- 

 ing the turns of the helix very close together. In this way, and by the use of 

 improved tungsten filaments that do not sag so readily at high temperatures, 

 and by using argon instead of nitrogen, it has gradually been possible to 

 constriict gas-filled lamps which are better than vacuum lamps down to 

 about 40 or 50 watts. These smaller lamps, although not much better in 

 efficiency than the vacuum lamp, have the advantage of giving a much 

 whiter light. In the case of the larger lamps, the use of the gas filling, 

 together with the special construction of the lamp, more than doubles the 

 efficiency. 



The invention of the gas-filled lamp is thus nearly a direct result of 

 experiments made for the purpose of studying atomic hydrogen. I had no 

 other object in view when I first heated tungsten filaments in gases at 

 atmospheric pressure. Even at the time that I made these experiments at 

 higher pressures, they would have seemed to me useless if my prime object 

 had been to improve the tungsten lamp. 



I hope I have made clear the important role that properly encouraged 

 scientific curiosity can have in industrial research. The illustration that I 

 have given is not at all exceptional. I could have given any one of several 

 others equally well. Many industrial laboratories have followed Doctor 

 Whitney's lead, in devoting a fairly large fraction of their activities to 

 these rather purely scientific researches. Certain men, at least, are not 

 expected to be responsible for practical applications, but are freely allowed 

 to make fundamental scientific investigations. The type of man who does 

 this work best can usually only be attracted to those industrial laboratories 

 that have adopted this policy. 



However, I do not believe that this second method of research is grow- 

 ing in popularity solely because it is found to be profitable. I feel rather that 

 most of our leaders in industrial research are eager to adopt this method, 

 in so far as economic factors may permit, because they realize the debt 

 that modern industry owes to the pure science of the past and because the 

 modern conception of service and the growing esprit de corps of American 

 industry help make them glad of any opportunity to contribute to scientific 

 knowledge. I know personally that such motives as these have guided 

 Doctor Whitney in the leadership he has taken. 



I believe in the near future there will be a much increased demand for 

 men with scientific training who are capable of doing more independent 

 thinking. 



