428 



THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1956 



With the recent rapid increase in operator dialed intertoll groups, it 

 might be expected that the above discrepancy between " % NC encoun- 

 tered" and "% NC existing" would disappear — for an operator now 

 initiates each call unaware of the momentary state of the load on any 

 particular intertoll group. By the use of peg count meters (which count 

 calls offered) and overflow call counters, this change has in fact been 

 observed to occiu'. ]\Ioreo^'er, since the initial re-trial intervals are com- 

 monly fairly short (30 seconds) subsequent attempts tend to find some 

 of the previous congestion still existing, so that the ratio of overflow to 

 peg count readings now exceeds slightly the "% NC existing." This 

 situation is illustrated in Fig. 5, which shows data taken on an operator- 



1.0 



AVERAGE SUBMITTED LOAD 



Fig. 4 — Observed proportions of time all trunks were busy on Albany and 

 Buffalo groups of 2, 4, 5, 7, and 9 trunks, 



