546 



BELL SYSTEM TECHNICAL JOURNAL 



ing times for dififerent group sizes at different levels of usage and compared 

 with delay intervals developed from the formula. This comparison is 



Comparison of 1945 Study with Cleveland Study of 1929-30 



Based on 3.5 Min. HT per Circuit Attempt 



or 5.25 Min. HT per Message 



The minutes of circuit delay shown above for the Cleveland study are derived by 

 subtracting the minimum speed of 1.65 minutes from the actual overall speed for the 

 various sizes of groups and levels of usage. The comjjarable 1945 figures are taken from 

 Fig. 5. 



It will be noted that the principal differences occur on the smaller groups at the higher 

 levels of usage. This is undoubtedl\- due to the fact that the alternate routes are more 

 heavily loaded today than they were in 1929-30 and therefore are less helpful in absorbing 

 the overflow from the first route. 



The holding time used in this comparison as j^roliabl} typical of 1929-30 is derived as 

 follows: 



Conversation time 3 .OOminutes 



Operating time 2.25 minutes 



5.25 minutes 



No. of Circuit Uses per Message 1 • 50 



HT per Circuit Use (5.25 H- 1.50) 3.50 



Fig. 4 



shown in Fig. 4. It will be seen that there is substantial agreement between 

 the two sets of delay factors, such differences as there are being explained 

 in the notes on that figure. 



