TIIJO STRUCTUUK OF THE NUCLEUS. 1 05 



battery. I have foiiud it more convenient, however, to keep F, = 0, whence 

 = aV, V^ - a V;/2 for small deflections. 



If the mean deflection be taken (since the constant aV:/2 mei'ely determines 

 the zero), d = aV,^V^-a V;/2 and -d=- a V^ F,, - a V\/2, whence 

 d = aV^V^, where V^ is the full potential difference available. In practice the 

 positive and negative deflections are numerically equal, but the scale zei'o varies 



with Fg. 



Finally, if the instrument is used idiostatically and F, = 0, V,, = V , 

 6 = a Vi/2. In this case the instrument gave me a= .036. 



If ^\ =V^ — 0, the needle and one ipiadrant being put to earth. 

 d^(xV^{— F2/2) = — aVl/2, the same value but opposite in sio-n. 



3. Apparatus. — In the figure (sectional elevation) QQ will be lecognized as 

 the quadrants, A as the base plate of an ordinary Elliott lecture pattern of the 

 electi-ometer. Details are omitted. A is to be used as a receptacle for the desic- 

 cator and in the definite form of instrnment must be deeper and provided with a 

 central hole for convenient dischai-ge. The needle is shown at /m, and m is the 

 mirror attached to the vertical stem. This carries at its lower end a metallic disc, 

 c, all but fitting into the cylindrical cavity of the brass box, h. When the needle 

 is at rest, the disc, c, lies at right angles to the axis of this box, and the free space 

 between the edge of the disc and the inside of the box is about 1 mm. The hollow 

 cylinder is appropriately closed at both ends, forming a capsule and it is supported 

 by the hoi'izontal glass I'od gg, fastened down with bolts to the ends of the bed 

 plate, A. A sleeve, d, serves to join the ca[)sule to the rod. 



The needle, 11, and disc, c, are made of very light aluminum foil, .03 thick. 

 In order to secure the necessary longitudinal rigidity in the needle, the foil is 

 folded along the axis as shown at ^, and rolled down nearly flat. The needle so 

 obtained weighs with mirror less than .7 grams. Its [)erioil in the earlier experi- 

 ments was 30 sec, in the later experiments 60 sec. (remembering that but 1/2 

 swing is usually needed) and the dam|)ing ratios 8:1 and 37:1 respectively. 

 Originally my needles and discs were made of silver-plated mica, but I was unable 

 thus to get the same lightness and ligidity as with aluminum. In the final 

 experiments the disc was even made annular as the damping in case of nariowed 

 quadi'ants is already excessive. 



As a result of this extreme lightness lelative to surface, the air damping 

 obtained in disc and capsule may reach aperiodicity, while the torque of the bifilar 

 is correspondingly reduced. The lids of the box, h, ai'e adjustable, the i-ing aa 

 carrying a narrow horizontal bar, h, centrally perforated so as to admit the .screw, 

 /, the inner end of which cari-ies the discs, e, e, by which the capsule is closed. The 

 outer ends of /,/are provided with a nut, which in view of the springs, ,s>, may be 

 used to regulate the distance between e and h at pleasure. A simi)le lid, however, 

 was found to suffice equally well, and the discs, e, e, may be left as much as 2 cm. 

 apart, admitting of ample swing of the needle. In fact so small is the amount of 

 induction transmitted by the discs that they may be made of (jlim thi'ough which 

 the edge of the vane c may be seen and its fi-eedom tested. The vane, c, is received 



