February 8, 1889.] 



SCIENCE. 



97 



albo-carbon lig'ht is burned in clusters of burners, a much more 

 striking effect, with a far higher resuh in lighting-power, is pro- 

 duced. An eight-light cluster tested on London i6-candle gas 

 gave 8.84 candles per cubic foot, while with the larger clusters the 

 increase in candle-power is still greater. This light may be seen in 

 use every evening at 728 Broadway, this city, the office of the Albo- 

 Carbon Light Company. 



THE NEW WESTON VOLTMETER. 



It is a law of human progress, nowhere better exemplified than 

 in the industrial and mechanic arts, that all systematic and perma- 

 nent advance depends upon our ability to determine quantitatively, 

 in terms of some standard, the value of the various factors involved 

 in any given operation or transformation. An idea of the crude- 

 ness of men's notions of measurement in former times as compared 

 with the present, may be gained from the names of units or stand- 



sitates the elimination of guesswork and the substitution of knowl- 

 edge. Engines, dynamos, batteries, electric motors and lamps, 

 are sold with a guaranteed efficiency and life, subject, however, to 

 definite conditions as to use. It is generally because of the absence 

 of definite knowledge as to when the imposed conditions are 

 actually observed, that losses so frequently exceed profits. When 

 measurements of the value of electrical appliances are actually 

 made, the results are often discredited because of doubt as to the 

 accuracy of the instruments used, and probably the general indiffer- 

 ence to accurate work manifested by many electricians may be justly 

 ascribed to the absence of reliable measuring instruments. 



Most of the commercial electrical measuring instruments in use 

 in this country to-day are of foreign manufacture. They may be 

 briefly described as either of the permanent or electro magnet type. 

 The former are in general disfavor, and unjustly so, because their 

 defects are not inherent in the types, but are, rather, the natural 

 result of poor design and construction. Those who have had much 



ards which have been handed down to us. Thus three barley- 

 corns made one inch, the foot was the length of the king's pedal 

 extremity, the hand is a measure still in use in estimating the 

 height of horses, etc. Compare such notions with the accuracy 

 required in modern machine-shop practice. It is, in fact, only 

 when the value of work already done becomes known, that one is 

 prepared to make further progress, as every step in advance de- 

 mands increased refinement in the means and methods of measure- 

 ment. As an instance in point, witness the mutual development of 

 the steam-engine and the steam-engine indicator. Just as the 

 indicator has advanced to a state of perfection such that its records 

 are universally relied upon to detect faults in present apparatus, 

 and intelligently outline the direction of improvement, so has there 

 been a gradual advance in the construction of commercial electrical 

 measuring instruments, serving a similar purpose, and tending to 

 effect a similar result, in electrical engineering. 



While it is true that thousands of engines are never indicated, 

 and thousands of electrical appliances are never carefully tested, it 

 is equally true, as a consequence, that useful energy is wasted, prop- 

 erty destroyed, and money lost. 



In electrical as well as in mechanical engineering, success neces- 



experience with these instruments, of which the Deprez and Ayrton 

 & Perry are examples, will recognize the fact that while these in- 

 struments are new, they are subject to rapid and serious changes 

 in their constant. The rate of change, however, diminishes with 

 age and use, up to a certain period, when they assume a condition 

 of stability, and are thenceforth, in so far as the controlling force 

 is concerned, reasonably reliable. Uncertainty as to when this con- 

 dition of stability is attained necessitates frequent calibration, and 

 is thus a serious obstacle to accurate work. A second defect is the 

 heating error introduced when the instruments were kept in circuit 

 even for the short time necessary to make readings. 



Instruments of the electro-magnet type are, on the other hand, 

 more generally in demand, because of the prevalent idea that they 

 are not subject to errors arising from a variable controlling force. 

 Errors, however, fully as serious as have been ascribed to perma- 

 nent magnets, are not only common, but seemingly inherent, in 

 this type, because of the magnetic persistency of the softest iron. 

 even when subdivided. This error is most noticeable when read- 

 ings are made with a rising, succeeded by a falling current, and 

 often amounts to as much as twenty-five or thirty per cent. The 

 best forms of this type of instrument are, perhaps, the ingenious 



