R.\nioHioi.oc;ic:Ai. mi.ciiamsm ai iiii: clllui.ak liakl 



forming network, a second identical network was prf)vided which could be 

 trickle charged from the same 20 k\' power pack and discharged indepen- 

 dently through the pulse transformer by means of a separate hydrogen 

 thyratron. Trigger pulses at the desired spacing applied to the two hydrogen 

 thyratrons then give twin output pulses of identical size and form, but var- 

 iable spacing. The changeover from repetitive pulses to single or twin pulses 

 is efTected by a single switch on the master oscillator panel, and this rajjid 

 changeover is very convenient since it is usually necessary to run briefly on 

 repetitive pulses in order to degas the guide immediately prior to single or 

 twin pulses. During this degassing the gun filament is off and the gun itself is 

 disconnected from the pulse transformer by a servo operated switch in order 

 to eliminate dark current. 



U 6 



Depth mm 



Figure 3. Depth dose for 1-4 MeV electrons in 

 Perspex 



[Reproduced by kind permission of C. W. Miller and J. W. Boa?) 



Any chosen number of pulses between 2 and 99 can also be delivered by 

 means of a twin-Dekatron pulse selector which stops the master oscillator 

 when the chosen number of pulses has been delivered. 



A magnetic holding field is provided by low voltage high current solenoids 

 surrounding the accelerating structure. These provide an axial magnetic 

 field which is 600 gauss near the input and tapers to a low value at the output 

 end of the machine. With this arrangement 50 per cent of the total electron 

 current ofO- 7 A (at 1-1 MeV) emerges through a thin aluminium window 

 within an 8 mm diameter circle. Scattering in the window and in the air 

 spreads the beam rapidly and the radial current distribution at any distance 

 Z along the axis may be represented very closely by a Gaussian curve. 

 Measurements of the radial distribution of dose rate at different distances 

 from the window have been made by irradiating small pieces of Perspex 

 arranged across the beam, and measuring the induced u.v. absorption rela- 

 tive to a control. This technique has been calibrated calorimetrically-^ and 

 yields rapid and reproducible estimates of dose. A typical depth dose curve 



290 



