378 Reports on Special Researches 



fact that it is to the radioactive material in the atmosphere and to the cause which is respon- 

 sible for the production of ions in a closed vessel (the so-called penetrating radiation) that we 

 must look mainly for an explanation of the normal atmospheric ionization. A further inter- 

 est attaching to the measurements of the number of pairs of ions produced in a closed vessel 

 arises from the fact that the formation of such ions has always been more or less a mystery, 

 which, in the case of land observations, is in part to be explained by the 7-ray radiation from 

 the radioactive materials in the soil, a cause which has very Uttle counterpart over the 

 ocean. 



In one sense, atmospheric-electric observations over the ocean are susceptible of a more 

 uniform interpretation than is the case with those taken on land, for at sea we are not 

 troubled with topographical features which vary from place to place. 



INSTRUMENTAL APPLIANCES. 



Among the chief difficulties associated with atmospheric-electric work at sea is that of 

 overcoming the effect, on the instruments, of the motion of the ship, and of securing good 

 insulation. One is practically debarred from the use of instruments of the quadrant type 

 and is forced to confine himself to electroscopes. The electroscopes in use on the Carnegie 

 are of two types, the bifilar electroscope designed by Wulf ,' and the single-fiber electroscope 

 of Einthoven, modified according to the Wulf pattern.- In each of these instruments the 

 restormg force, which resists the motion of the fibers or fiber under the action of the 

 electrical forces, is brought about by the tension of a quartz bow, so that the indications 

 of the instruments are affected to a comparatively small extent by the motion of the ship. 



It will be recalled that in the bifilar instrument the gold leaves of the older forms of 

 electroscope are replaced by platinized quartz fibers. The fibers are soldered at their 

 upper ends to the main terminal of the instrument and at their lower ends to the mid-point 

 of a quartz bow whose ends are fixed to a frame. ^Vhen the fibers are charged they repel 

 each other, and the resultmg motion, which can be read by a microscope with a scale in the 

 eyepiece, is resisted by the quartz bow. This type of instrument is useful where a sensi- 

 tivity in the neighborhood of 0.5 division per volt is requu-ed. Further, the case of the 

 instrument is double, and the inner part is insulated, so that by raising or lowering its 

 potential, by means of batteries, the readings of the electroscope can always be brought to 

 the most uniform part of the scale. The subsidiary case has an additional advantage in 

 enabling the electroscope to be used for any desired range of potential. 



In the single-fiber electroscope, a single platinized quartz fiber is attached at its lower 

 end to a quartz bow and at its upper end to the main terminal of the instrument. Two 

 insulated metal plates are mounted with then planes parallel to each other and to the 

 quartz fiber, one plate being mounted on each side of the fiber. The case of the instrument 

 being earthed, these plates may be charged to say + 100 volts and - 100 volts respectively, 

 or to any convenient amount, by means of constant batteries, and charges communicated 

 to the fiber will then cause a deflection. The deflection for a given potential applied to the 

 fiber increases with the field between the plates and with diminution of tension on the fiber, 

 which latter may be varied by moving the bow support up and down by means of a suitable 

 screw. In the laboratory it is not difficult to secure a sensitivity of 100 or more eyepiece 

 divisions per volt, but on board ship a sensitivity of from 5 to 10 divisions per volt is 

 found more desirable. 



For the batteries which determine the potentials of the plates of the Einthoven instru- 

 ment, small groups of cadmium cells (Kruger batteries) are generally recommended. 

 Unfortunately these batteries are liable to show sudden fluctuations in voltage, which, 

 though of small amount, are sufficient to cause erratic movements of the fiber of the electro- 

 scope. The resistance of a 100-volt Kruger battery is very high, about 100,000 ohms, and 



^Phya. Zeit., vol. 8, pp. 246 and 527, 1907. 'Phys. Zeit.. vol. 15, pp. 250-254, 1914. 



