560 



BELL SYSTEM TECHNICAL JOURNAL 



The operators did not make NC notations on the tickets in a certain 

 proportion of the cases where NC was actually encountered. 



Because the operator does not test for an idle trunk with machine- 

 like finality there are probably many cases where she does not con- 

 sider that an NC condition existed if it was of such short duration 

 that it did not materially affect her ability to secure a trunk. 



The possibility of some "limited sources" effect in the case of small 

 groups. The number of people in Newark who have occasion to 

 call York, Pa. must be relatively small since only one trunk is pro- 

 vided. Therefore, while the trunk is in use on one call there is less 

 likelihood that a second call will be originated than would be the 

 case if there were a greater community of interest between the two 

 places. Although the NC condition exists during the period of one 

 call, no NC is encountered unless there is a second and overlapping 

 call. 



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 



SIZE OF GROUP 



Fig. 13— Per cent, of calls delayed by NC (with alternate routes where authorized). 



Because of the importance of Items 2 and 3 above, both of which involve 

 the testing of trunks by operators, the empirical data should be used as 

 representative of per cent. NC encountered with ringdown operation 

 (operator testing) and the Erlang "B" per cent. NC existing should be used 

 as representative of intertoll diaUng conditions (mechanical testing). 



