678 



SCIENCE 



[N. S. Vol. XXXIII. No. 853 



men. In the case of our metallized carbon 

 filament, which has now been in use several 

 years, the efficiency of the light was in- 

 creased by about twenty per cent. Among 

 the carbon lamps of last year these were 

 sold to the extent of over a million dollars. 



A broader, but perhaps less accurate 

 impression of changes recently produced, 

 may be gained by considering the economy 

 now possible on the basis of our present in- 

 candescent lamp purchases in this country 

 and that which would have resulted if the 

 lamps of only ten years ago were used in 

 their stead. On the assumption that the 

 present rate of lamp consumption is equiv- 

 alent to about eighty million 25 watt tung- 

 sten lamps per year, and on the basis of 

 one and a quarter watts per candle power 

 as against 3.1 of the earlier lamps and 

 charging power at 10 cents per kilowatt 

 hour, we get as a result a saving of $240,- 

 000,000 per year, or two thirds of a million 

 per day. Naturally, this is a saving which 

 is to be distributed among producers, con- 

 sumers and others, but illustrates very 

 well the possibilities. It is interesting to 

 note that we are still very far removed 

 from a perfect incandescent illuminant, 

 when considered from the point of view of 

 maximum theoretical light efficiency. 



I see from advertisements that 65,000 of 

 the magnetite arc lamps, originally a prod- 

 uct of the laboratory, are now in use. 

 These must have been sold for something 

 near $2,000,000. The supplying of elec- 

 trodes, which we make and which are con- 

 sumed in these lamps, should amount to 

 about $60,000 per year. 



Our study of the properties of the mer- 

 cury arc produced our rectifier, which has 

 been commercially developed within the 

 past few years. Of these, about 6,000 have 

 been sold. As they sell for not far from 

 $200 per set, it is safe to say that this also 

 represents a sale of over a million dollars. 



The advantage of these outfits over other 

 available apparatus must also be recog- 

 nized as not far from $200 for each hour 

 through which those already sold are all 

 operating. 



In such a complex field as insulations 

 and molded materials there have been 

 many changes produced. As far back as 

 1906 we were using annually, in a certain 

 apparatus, 30,000 specially drilled and 

 machined soapstone plates, which cost 

 $1.10 each. As the result of experiments 

 on substitutes for such material, it was 

 found that they could be molded by us in 

 the proper shape, with holes in place and 

 of a material giving increased toughness, 

 at a greatly reduced cost. As the result of 

 this fact, the price of the purchased mater- 

 ial was reduced to us from $1.10 to 60 

 cents which in itself would have paid for 

 the work. But further developments 

 proved that the new molded material could 

 be made for 30 cents, which the foreign 

 material could not equal, so we have since 

 produced it ourselves. This caused a sav- 

 ing of approximately $24,000 annually for 

 this one molded piece. I have heard of 

 other cases where prices to us have gone 

 down, when we have obtained a little 

 promise from our experimental researches. 



In considering the research laboratory 

 as a financial asset there is another view 

 which might not be visible at first sight. It 

 is the question of the difference between the 

 value of the useful discovery when pur- 

 chased from competitors in the business 

 and when made by one's own company. 

 It is not usually pleasant to have to pur- 

 chase inventions after their value is known, 

 no matter from whom, but to have to pay 

 a competitor for such a discovery is 

 doubly irksome. One is naturally unduly 

 fearful of its value to the competitor, and 

 he, in turn, is overestimating another's 

 power to use it. The purchaser's profit is 



