620 ADOLF F. VOIGT 



inside a glass shell. The center wire, usually tungsten, is supported 

 on one end only and has a small glass bead on the free end to prevent 

 sparking. The window affixed to a metal or glass flange on the open 

 end of the cylindrical tube is of high grade mica split to the desired 

 thickness, which can be anji;hing from 10 m up- The minimum 

 allowable thickness is that which will support a vacuum and will 

 depend on the area of the window. The window must be of uniform 

 thickness and it will be if split along cleavage planes. If not uniform, 

 lines marking a sharp variation in thickness will be visible. 



End wdndow counters are available from a number of commercial 

 sources {84-87), and additional manufacturers are coming into the 

 market at the present time. So many different types of construction 

 are available that various recent catalogs should be consulted. These 

 tubes are expensive, particularly so since their life is limited and the 

 windows are fragile. Refilling services are offered by the manu- 

 facturers, but there may be some question as to the value of refilling 

 a tube with fresh gas after the quenching gas has been decomposed, 

 since other effects such as pitting of the anode may also occur. Tubes 

 similar to these can be machined and built in the laboratory, but there 

 is a considerable art to the making of good tubes, which needs to be 

 acquired before any consistent success can be expected. A descrip- 

 tion of such a tube is given by Copp and Greenberg {30). It is not 

 expected that the small laboratory entering this field will have the 

 time or patience to develop the art. Although the commerical cost 

 may seem high, considerable experience is necessar}- before a reason- 

 able percentage of homemade tubes will operate successfully. (For 

 further discussion of Geiger tube construction and action see among 

 others the follo^A-ing references: i, p. 58; 2, pp. 105. 141; 28,31.) 



Counter Circuits and Scalers. Some way of recording the 

 counts must be available. The pulses from a Geiger counter come 

 in much too rapidly and are too randomly spaced to be recorded 

 efficiently by a mechanical counter, however rapid its action. For 

 this purpose an electronic circuit known as a "scaler" is always used. 

 Such a circuit is composed of a number of units, each of which 

 will reduce or scale down the number of pulses reaching it by a factor 

 of two. A "scale of two" will record the first pulse reaching it by 

 lighting a neon glow lamp. In receiving the second pulse the glow 

 lamp is extinguished and a pulse is sent on to the next scaHng stage. 

 The fraction of reduction in counts ^nll be equal to 2 raised to a power, 

 that power being the number of scales of two in the circuit. The 

 most commonly used circuits employ sLx such scaling stages and are, 



