10 



OCEAN ATMOSPHERIC-ELECTRIC RESULTS 



motor used previously, and put in good repair, was in- 

 stalled for cruise VH. A description of the apparatus, 

 the procedure for making measurements, and the method 

 of computing ion content was given Ijy Swann in 1917 (8). 

 Some details of both apparatus and observational pro- 

 cedure, however, can be described to advantage at this 

 time. The apparatus differed from the design of Ebert 

 in one major respect; namely, it was designed so that 

 the "charging" method instead of the "discharging" 

 method could be used. In the Ebert apparatus a stream 

 of air is drawn by a fan through a cylindrical condenser, 

 the inner cylinder of which is connected to an electrom- 

 eter and charged to a potential sufficiently great to in- 

 sure that all small ions (mobility greater than 1.0 cm per 

 second per volt per cm) are removed from the air stream 

 passing by it. Ions of sign opposite to that of the charge 

 on the inner cylinder will be collected by that cylinder, 

 discharging the cylinder with time. 



In the charging method the air is drawn through a 

 cylindrical type condenser, the outer cylinder of which 

 is maintained at a constant potential while the inner 

 cylinder is, at the outset of an observation, near to earth 

 potential. Ions which are well within this condenser and 

 of the same sign as the potential on the outer cylinder 

 are driven toward the inner cylinder. When the apparatus 

 is used as an ion counter, all ions of mobility greater than 

 a certain value reach this cylinder and impart a charge to 

 it, hence the designation "charging method." The in- 

 crement of charge thus acquired from a known volume 

 of air is measured and used in the calculation of ionic 

 density. Ions which at any instant are located outside or 

 even some distance within the mouth of the condenser, 

 however, are subjected to a field which opposes their 

 entrance unless special provision is made when design- 

 ing the condenser to shield them from an opposing field. 



Swann (9), who devised this type of counter, was cog- 

 nizant of this potential source of error. He gtates "this 

 difficulty is one which shows itself very materially in 

 practice, as appeared when experiments were made to 

 test it ... I finally adopted (the shielding cup) which en- 

 tirely overcomes the difficulty." The effectiveness of 

 this device has been verified by tests made on several 

 occasions in more recent years. 



A more detailed description than that given by Swann 

 of the novel features of the type of ion counter used in 

 the Department of Terrestrial Magnetism is given in 

 that which follows. Some improvements which were 

 made when conditioning the instrument for use on the 

 Carnegie during cruise VII are included. 



This type of counter differs from the well-known 

 Ebert only in the form of condenser and in that a single - 

 fiber electrometer is used. It will suffice here to de- 

 scribe only the condenser. The details are shown infig- 

 ure 23. It consists of an outer brass cylinder (1) with 

 its attached cap (9) and intake-port (10) which is covered 

 with a suitable cowl (not shown), an inner cylinder (2), 

 and a central electrode or collector which is made up of 

 the brass rod (4), the struts (7), and the cylindrical 

 brass shield (3). The lower end of the outer cylinder 

 fits into the cap of the electrometer and connects with 

 the tube that leads to the air -flow meter and the air- 

 flow fan and is thus in electrical contact with all the 

 "earthed" parts of the system. The inner cylinder is 

 insulated from the outer cylinder by two tight-fitting 

 ebonite rings (5), which, together with the binding-post 

 (6), also serve to support it. This cylinder is maintained 

 at a constant potential of from 20 to 100 volts by means 



of a battery attached at the binding-post (6). The lower 

 end of the central rod (4) fits into the post of the elec- 

 trometer and thus connects with the sensitive element. 

 Three brass struts (7) (radial arms also have been used) 

 support the thin-walled brass shield (3) from the central 

 rod and connect these parts electrically. The shield (3) 

 is maintained in a central position by an amber ring (8). 

 An annular disc (11) is placed below this insulator to 

 shield it from the intense field of the condenser -ele- 

 ment (2) and thus avoids effects from induced creeping 

 charges. The shield (3) is designed to intercept those 

 electrical fields which, without it, would dfive some of 

 the ions to "earthed" parts of the condenser. Thus, 

 for example, the field which without (3) would extend be- 

 tween the upper part of the inner cylinder (2) and the 

 lower part of the intake-port (10) and cap (9) would also 

 drive some ions to these last-mentioned partsandthere- 

 by decrease the number collected by the central elec- 

 trode. The number of ions determined under these con- 

 ditions accordingly would be too small. 



The relative merits of this type of ion counter and 

 the usual Ebert type for conditions encountered on the 

 Carnegie at sea are as follows: (1) An increment of 

 charge may be measured with a given precision in about 

 one-twentieth the time required for the Ebert, provided 

 that in both methods the conductance across the prime 

 insulators is small and nearly constant. (2) Defective 

 insulation is both a less common and a less serious 

 source of error than with the Ebert method. (3) With the 

 charging method one can at once distinguish corrections 

 arising from insulation -leak from those arising from a 

 conduction current. (4) With the charging method more 

 auxiliary equipment is needed than with the Ebert meth- 

 od, but this disadvantage is not serious and is greatly 

 outweighed by the advantages enumerated. 



Ion counter 1 was mounted in the atmospheric-electric 

 cabin near the conductivity apparatus (fig. 3). It was held 

 in a gimbal supported from the ceiling, and the air -flow 

 tube projected approximately one -quarter meter through 

 a roof aperture (fig. 8). A funnel was fitted at the intake 

 which could be turned into the wind and thus eliminate 

 possibility of aspiration up the air-flow tube in moderate 

 and heavy winds. A domed metal cap or hood covered the 

 roof aperture and air-flow tube when the apparatus was 

 not in use. The potential applied across the cylinders of 

 the condenser on cruise VII, was 60 volts from May 1, 

 1928 to July 28, 1929, and 90 volts thereafter. 



The unifilar electrometer was adjusted for a sensi- 

 tivity of approximately ten divisions per volt. The air- 

 flow fan was driven by an electric motor operated from 

 the ship's power plant, and the volume of air passed 

 through the apparatus during a period of observation was 

 determined by an anemometer calibrated by the United 

 States Bureau of Standards in March 1928, just before 

 cruise VU began. The same anemometer was used on 

 cruise VI and was calibrated by the Bureau October 6, 

 1919, before that cruise began and again on February 24, 

 1923, sixteen months after it ended. The correction fac- 

 tors for converting anemometer readings to liters per 

 second obtained on these three occasions were: for 1919, 

 1.08; for 1923, 1.05; for 1928, 1.02, over a range of 1.4 

 to 2.0 liters per second. The factor for cruise VII, 

 namely 1.02, was thus in reasonable agreement with the 

 values for the previous cruise. That it varied little, if 

 any, from the value of 1.02 during the nineteen months 

 of cruise VU seems likely, judging from the small change 

 indicated for the twenty-five months of cruise VI. 



