502 



NATURE 



[January 15, 1920 



of resistances and batteries. Considering the 

 anode circuit of the first bulb, we see that it con- 

 tains a resistance Rj and a high-voltage battery 

 Ej. Let Ej = 8o volts, R, = 30,000 ohms, and the 

 current be i millianipere. At present ignore the bat- 

 teries marked ej, Cj' ; then the fall of potential 

 along Rj is 30,000 x io~' = 30 volts. Such a 

 potential difference applied between the grid and 

 filament of the second triode would put this tube 

 completely eut of action. It is therefore neces- 

 sary to introduce a neutralising battery of about 

 30 volts at the point marked Cj, or at the point 

 marked Cj'. In the latter case the battery will, 

 in fact, be a portion of the battery E. Suppose 



Fig. 2 



this to be done, and imagine an electro-motive 

 force eg to be applied between the input terminals 

 of the amplifier. Then it can be shown that the 

 consequent increase of current in the anode cir- 

 cuit is 



,R,-f i//4»' 



where g is the triode 's voltage factor, and ha 

 its differential conductance. These parameters 

 are frequently of the order 



S=\o and //„=io"^^. 



The electro-motive force handed on to the second 

 triode from the terminals of Rj 

 is clearly of magnitude 



R. + i/V 

 Using the value of Rj suggested 

 above, we find that the multi- 

 plier of eg becomes 

 30X lo^x io_ 

 4oxio» ''^' 



The amplification approaches the limit 10 (that is, 

 the value of g) the greater we take the value of 

 Ri ; but obvious practical reasons limit the magni- 

 tude of this resistance. 



This type of amplifier, usually spoken of as a 

 resistance cascade amplifier, has been much used 

 by the Americans and the French for amplifying 

 rapidly varying currents, but in that case a con- 

 denser is substituted for the battery at ej; a "grid 

 leak " of order a megohm must then be connected 

 between grid and filament in each triode in order 

 to avoid the accumulation of negative electricity 

 NO. 2620, VOL. 104] 



on the now insulated grid. The complete instru- 

 ment is then usually connected so as to utilise a 

 common battery of 4 or 6 volts for all filaments, 

 and a common battery of about 80 volts for all 

 anodes. Adopting common battery connections, a 

 finished two-stage amplifier is seen in I'ig. 2. The 

 grid leak S connected between the first grid and its 

 filament might be about a megohm, and is neces- 

 sary only when the circuits from which the input 

 to the amplifier comes are such as would other- 

 wise leave the grid insulated. According to the 

 computations given above for one stage, the 

 amplification with this two-stage instrument 

 should be 



7-5 X 7- 5 = 56-fold. 



Finally, a mode of connection due to the present 

 writer may be described. It is shown in Fig. 3, 

 arranged to constitute a two-stage amplifier suit- 

 able for use with slowly varying currents. The 

 first triode of the pair has resistances Qj, Rj, Sj, 

 connected so as to constitute with the bulb the 

 four arms of a balanced Wheatstone bridge. The 

 high-voltage battery Ej is connected across two 

 opposite corners of the bridge. An electrical 

 stimulus applied to the grid causes the balance 

 to be disturbed, and a corresponding potential 

 difference arises between the filament lead and 

 the junction of Qj and Rj. This difference of 

 potential is conveyed to the next triode by direct 

 connections. The magnification is the same as 

 that obtained with the arrangement of Fig. 2 when 

 resistances Sj, S, are taken of the same value as 

 Rj in that figure. A grid leak S is used when 

 necessary, for the reason explained before. 



In conclusion, it would be well to point out 

 that in deciding upon the resistances and the 

 voltages to be employed in constructing these 

 amplifiers, the characteristics of the triodes should 

 be kept in mind with the view of using them all 

 at such adjustments that the relation between input 



Fig. 3. 



and output potential differences is linear, and the 

 magnification therefore free from distortion. 



VV. H. EccLES. 



INDIAN GEOLOGY.^ 



"T^HE appearance of a manual of Indian geology 

 -*- of so excellent a character as the present 

 work, written by an Indian geologist, is an event 

 of some importance, since it furnishes a fresh 

 and convincing answer to the argument so often 



» "Geology of India for Students. By D. N. Wadia. Pp. xx+^gH+xx 

 plates. (London : Macmillan .ind Co. Ltd., 1919. Price iBs. net. 



