58 



THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1952 



presented, namely for 0, 2, 5, 10 and 15 seconds of dial tone delays from 

 the instant of sender seizure. It should be noted that the general con- 

 tour of the various curves up to the receipt of dial tone and when ex- 

 tended beyond gives an estimate of the survivor curve for dial tone 

 delayed indefinitely. Fig. 14 shows the results for the flat rate class of 

 service (the data here are relatively meager), and Fig. 15 shows the 

 results for coin customers. 



Fig. 16 indicates for the three classes of service the progressive changes, 



1 31Alli Aa NOIiOV N3><Vi ION 3AVH OHM 58391805802 dO iN30 a3d 



