52 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1952 



such comparisons are shown on Figs. 10 and 11. The agreement is quite 

 good in most cases. 



SUBSCRIBER DIALING HABITS AS OBSERVED WITH A MONITORING CIRCUIT 

 ON A SENDER WITH INDUCED DIAL TONE DELAYS^ 



As a separate study a series of tests was made by means of a moni- 

 toring circuit on one of the senders serving in common the subscribers 

 in the Sterhng-3 and Main 2 central offices, for the purpose of obtaining 

 further information on subscriber diahng characteristics under overload 

 conditions. A large amount of data was collected on the time intervals 

 from the seizure of the sender to the first action taken by subscribers 

 when encountering dial tone delays, the latter being introduced under 

 the control of the observer. 



The monitoring circuit was wired to a particular sender in a group 

 of 100 serving all classes of subscribers. When the circuit was in use, the 

 only irregularity introduced was that the dial tone could be delayed 

 even though the sender was actually available to the subscriber. The 

 delay did not affect the sender in its functions if the subscriber elected 

 to dial before tone. 



The sender monitoring circuit provided the following four features: 



1. A receiver was bridged across the tip and ring leads in the sender 

 so that an observer could hear certain actions taken by a subscriber 

 connected to the sender. The sender was of course disconnected before 

 conversation. 



2. The observer was able to preselect one of several intervals by 

 which dial tone was delayed on successive calls served by the sender. 

 This was accomplished with a capacitance-resistance-vacuum tube cir- 

 cuit. 



3. By means of a timer which started when the sender was seized, 

 the observer was enabled to note elapsed time intervals to the occurrence 

 of the various actions of the subscribers. The reading of the time of 

 the first action of a subscriber had to be made when the second hand 

 was in motion, which introduced certain errors later to be discussed. 



4. By means of colored lamps the observer was able to classify all 

 calls observed as being message rate, flat rate or coin. 



During the sender dial tone delay tests, observations were made only 

 during the afternoons when the flow of traffic was light and the prob- 

 ability of a subscriber obtaining a delay before reaching the sender was 

 a minimum. 



* Based on an unpublished report by W. A. Reenstra. 



