DEPARTMENT OF TERRESTRIAL MAGNETISM. 259 



total circuit resistance) to an order of magnitude determined by the measuring 

 instruments employed. This may be done systematically if the individual 

 contact resistances are known. 



With the circuit resistance thus adjusted, the contact resistance has further 

 interest in so far as its variation indicates such physico-chemical changes in 

 the vicinity of the electrodes as would bring with them changes in the contact 

 potentials, and consequently false indications on the records. Thus, in 

 addition to insuring more suitable resistances when the electrodes are being 

 installed, this method will, it is believed, also furnish information helpful in 

 dealing with the troublesome spurious potentials which from time to time 

 arise at the electrodes. 



The essential to the method is to have two electrodes in the vicinity of each 

 terminal of the line (one of each pair may be only temporary) , so arranged that 

 they may be connected to the line singly or in pairs. Then if x and v are the 

 contact resistance of two electrodes at one end, and Ri the resistance measured 

 when X alone of the pair is in circuit, ^^2 that when x and v are both connected, 

 and Rz that when v alone is connected, then: 



x = R2-R,+ V{R2-Ri) {R2-R-iy 



The other resistances may be determined in a similar manner, from two 

 additional measurements. 



A technique for preparing the moving elements for string electrometers and Einthoven 

 galvanometers. O. H. Gish. 



The quartz fibers which are used for string-electrometer elements are made 

 by (a) "blowing out" quartz glass in an oxyhydrogen blast (see pp. 260-261). 

 They then (b) receive a coat of gold by the "electric spray" or "sputtering" 

 method, after which they are (c) tested for resistance, and fibers of the desired 

 resistance and size selected (d) to mount into a transferable element. 



The present report has little new to add under (a) and (b), except that 

 "kinks" may be removed from raw fibers by passing them through a low 

 Bunsen flame and that considerable time may be saved by "sputtering" 12 

 to 15 fibers at a time instead of 1 or 2, as has been done heretofore. The 

 latter is done by mounting the fibers on a small glass frame which is about 0.5 

 inch wide and 6 inches long, and which, when filled with fibers, has the 

 appearance of a miniature harp. 



The operations under (c) immediately precede the mounting of fibers into 

 electrometer elements. The first of these is to select a fiber (or in case of the 

 bifilar elements, two fibers) of the desired diameter, using for this purpose a 

 compound microscope of about 400 diameters magnification. The fiber is 

 then removed from the frame with a special tool which resembles a pair of 

 dividers. The chance of breaking fibers has been greatly lessened by pro- 

 viding this with flexible points made from strips of phosphor-bronze sheet 

 about 0.08 mm. thick. Fused shellac is used for attaching the fiber. It is 

 then brought into contact with two thin silver wires which extend so as to 

 form a pair of very flexible and parallel cantilevers about 10 cm. apart, and 

 the resistance measured with a Wheatstone bridge. The average resistance, 

 which of course varies with a number of factors, is about 300 ohms per centi- 

 meter length of fiber. 



In all mounting operations (d) where unmounted fibers are to be transferred, 

 a universal-motion device having both rough and fine adjustment is helpful. 

 Three such units are needed for the device to assemble the element for a 

 bifilar (Wulf) electrometer. Such a "mounting device" was designed and 

 constructed. With it much of the uncertainty in the mounting of fibers has 

 been eliminated, the quality has been improved, and the nervous strain on the 

 manipulator greatly reduced. 



