June 24, 1904.] 



SCIENCE. 



947 



times more than in the comparisons of in- 

 candescent lamps of approximately equal 

 efficiency with one another. Moreover, in- 

 candescent lamps suitably prepared and 

 properly used are very permanent, and, 

 being cheap and portable, may be dupli- 

 cated and frequently tested. By keeping 

 one set of lamps as reference standards and 

 another as working standards, and burning 

 them at relatively low temperatures (that 

 is, at about four watts per candle) there is 

 good reason for believing that the average 

 value of a set of standards may be con- 

 tinued indefinitely. 



A considerable number of electric stand- 

 ards have been obtained from the Reichs- 

 anstalt, the ratio of the candle to the 

 Hefner unit being taken as 1 to .88. These 

 reference standards are, of course, only oc- 

 casionally used, and the mean of the value 

 of several 16-candle power lamps is taken 

 as the standard of the bureau. Exact 

 copies of these wiU be added from time 

 to time, so that of a change in any 

 lamp is detected it may be discarded with- 

 out impairing the completeness of the set. 

 The current and voltage employed in test- 

 ing lamps are measured by a potenti- 

 ometer, and can be maintained constant to 

 the hundredth of one per cent. Working 

 by the substitution method, it is possible to 

 make very accurate comparisons and thus 

 to secure very exact copies of the standards 

 of the bureau. The bureau recently re- 

 quested a large number of lamp manufac- 

 turers to send each two or three carefully 

 rated 16-candle power lamps for compari- 

 son with our standards. The lamps sub- 

 mitted varied from 15.4 to 17.6 candle 

 power, averaging 16.48 cp., or about three 

 per cent. high. Several of the large manu- 

 facturers were quite near to our standard, 

 and it appears from these results that if all 

 lamp manufacturers were to adopt the 

 standards of the bureau there would be very 

 little change in the average candle power 



of all the lamps manufactured, although 

 ■ some would be raised and others lowered. 

 Since uniformity is extremely desirable 

 and is now more easily attainable than here- 

 tofore, it is to be hoped that this result will 

 speedily follow. The close agreement be- 

 tween the standards of the bureau and 

 those of some of the manufacturers is due 

 to the fact that the latter are using incan- 

 descent lamps rated at the Reichsanstalt. 

 For the same reason, standards of the bu- 

 reau are also in close agreement with those 

 of the Lamp Testing Bureau of New York. 

 The purpose of the bureau is not to un- 

 dertake, at least for the present, the com- 

 mercial testing of incandescent lamps, but 

 (apart from the testing done for the gov- 

 ernment) only to verify lamps to be used 

 as standards and to make special investi- 

 gations of lamps submitted for the pur- 

 pose. To this end no effort will be spared 

 to maintain reliable standards and to 

 certify copies with the highest possible 

 precision. 



DIVISION III. 



As already stated, the chemical division 

 was late in being inaugurated. Aside from 

 the immense assistance which a chemical 

 laboratory can render to physical investi- 

 gations, the division of chemistry will have 

 important functions in its relations to the 

 chemical interests of the country, and to 

 the customs service and other departments 

 of the government. Some chemical work 

 is now being carried on, and detailed plans 

 are being developed for the chemical labo- 

 ratory to be installed in the larger of the 

 two buildings now under construction. 



THE EXPOSITION LABOEATOET. 



In addition to the exhibit which the bu- 

 reau is making in the government building 

 at St. Louis, it has undertaken, at the re- 

 quest of the authorities of the exposition, 

 to install and operate an electrical testing 

 laboratory in the electricity building dur- 



