February 8, 1889.] 



SCIENCE 



99 



Another very important feature is the means afforded those not 

 in possession of laboratory appliances of verifying the scale value 

 by applying a single cell of some constant form, such as is to be 

 found m any telegraph-office, to the terminals of the coil giving the 

 lower scale-reading. The deflection noted serves as a standard for 

 future comparison with the same or a similar cell, should doubt 

 arise as to the effect of accidental rough usage. 



All scale-readings begin at zero, and extend by practically uni- 

 form increments to the maximum reading. The range of scale- 

 readings for instruments of a given maximum scale is thus greater 

 than is common ; and, as the divisions of each scale are the result 

 of individual calibration and checking, the scale-readings are uni- 

 formly accurate. The temperature correction is negligible, and the 

 instruments can be kept constantly in circuit, as their resistance is 

 so high (averaging twenty thousand ohms) as to prevent any 

 appreciable healing error. The ammeters have the same general 

 appearance as the voltmeters, and possess the same merits of per- 

 manency and reliability. 



In the hands of electricians and electrical engineers, these instru- 

 ments are claimed to afford the means of obtaining measurements 



/^ Jill 



-correct to within one-fifth of one per cent, and special instruments 

 are made correct to within one-tenth of one per cent. If the limits 

 of error were even ten times as great as claimed, these instruments 

 would, it is said, possess greater accuracy than has been heretofore 

 attainable in commercial voltmeters and ammeters. It is most 

 certainly to be hoped that actual practice will substantiate the accu- 

 racy of these claims. 



ELECTRICAL NEWS. 

 Hertz's Researches on Electric Oscillations.' 



In order to get resonance phenomena between two circuits. Hertz 

 used an arrangement consisting of a straight copper wire divided 

 into two parts by a discharger, the two halves being connected with 

 the secondary of an induction-coil, while two hollow zinc spheres 

 were arranged to slide on the halves. The micrometer circuit was 

 made of such dimensions as to have a slightly shorter period than 

 •that of the discharge circuit, supposing the oscillations were really 

 as rapid as was calculated. The experiments were made in two 

 ways. First, the period of the micrometer circuit was increased : 

 the result was an increase in the length of the spark that could be 

 obtained in it, followed by a decrease, as the capacity, and therefore 

 the period, became too great. Afterwards, the micrometer circuit 

 remaining constant, the period of the discharge circuit was de- 

 creased, the result being, as before, an increase in spark-length in 

 the micrometer circuit, followed by a decrease. 



' Continued from No. 313. 



We may fairly conclude, then, from all of these experiments, that 

 the effects observed in the micrometer circuit were produced by 

 oscillations in the discharge circuit of a period approximately equal 

 to that calculated from the dimensions of the apparatus, in the 

 neighborhood of a hundred-millionth of a second. 



Hertz concluded, that, if vibrations were caused in the microm- 

 eter wire, there must be nodes (points of zero disturbance) somewhere 



; showing relation between length of side of rectangle (taken as abscissa) and 

 sparking distance (taken as ordinate), the sides consisting of straight 

 ; of varying lengths. 



along its length. To prove this, he adjusted his micrometer circuit 

 to resonance with the discharge circuit, making the gap in the 

 former so wide that sparks were just able to pass. Then a sphere 

 was made to touch different points along the wire, the result being 

 a cessation of the sparks except When the point of contact was at 

 the middle, showing that there was a node at that point. Again, 

 by using a second micrometer circuit similar to the first, as in Fig. 5 

 (Fig. 7 in the paper), nodes were found to occuronc^and^g"/?. When 

 the wire connecting 2 and 4 was removed, the vibrations were not 

 disturbed ; but when the knobs at these points were brought close 

 together, a slight spark was observed between them, the spark 

 corresponding to a vibration with a single node at ae. We can, 

 then, in the same conductor have vibrations with one or two nodes, 

 according a,s we wish ; that is, we can excite in it its fundamental 



Curve showing relation b 

 maximum sparking d 

 gradually drawn out. 



length ( 



rectangle (taken as abscissa) and 

 e), the sides consisting of spirals 



note or its first overtone. As to the higher overtones, Hertz con- 

 siders it doubtful whether it is possible to produce them, for the 

 results show that the damping effects must be considerable ; and 

 there are many secondary phenomena which show that irregular 

 vibrations are superposed on the regular ones. To obtain the best 

 results, Hertz observes that there is a longer spark in the secondary 

 when it is exposed to the light of the discharge circuit. 



Let us now call the discharge circuit the primary, and the microm- 

 etercircuit the secondary. The next experiment Hertz tried was with 

 a primary circuit of straight copper wire, carrying at its ends zinc 



