TiiU.NK KEQUIUEMEXTS lA ALTEKXATE liUUTlNG XETWOKKS 285 



value of 1.12. With the latter ratio the economic CC8 would be (28 -^- 

 1.12) or 25. On Curve A of Fig. 3 it will be seen that tiunk No. 5 (the 

 last trunk of a 5-trunk group) carries 25 CCS. So, if the lower cost ratio 

 had been used insteatl of the correct one, the effect upon overall trunking 

 costs would ha\e been 1 per cent in excess of the most (u-onomical ar- 

 rangement. Had a cost ratio of 1.25 ])een used, six trunks would have 

 been provided in the IIU group and no cost penalty would have been 

 incurred. On the other side of the optimum point an eight trunk HU 

 gi'oup would meet the requirements of a cost ratio as high as 2.0 with a 

 resulting cost penalty of about 2 per cent. As can be seen from Table I, 

 the cost penalties mount more rapidly when more than the optimum 

 number of high usage trunks are provided than when less are provided. 

 The principles of alternate routing and certain of the technitjues used 

 by traffic engineers in determining (luantities and arrangements of inter- 

 office trunks have just been described with particular reference to the 

 trials that have been cai'ried on in New York City. The latter were very 

 extensi^•e undertakings in which not only single alternate routes were 

 provided but for a majority of items, multiple alternate routes. This was 

 possible because New York City had two tandem systems each with a 

 completing field to all city offices as well as other tandem systems (office 

 selector tandems) each with a completing field to about 20 offices. Thus 

 it was possible in many cases for an originating office to test a direct 



Table I — Comparative Costs of Alternate Routing System 

 FOR Various Assumptions as to Number of HU Trunks 



Given: Offered load in CCS 240 



Efficiency of trunks added to alternate route 28 



Cost ratio, alternate to direct (HU) route 1.4 



* Overflow CCS from HU Group X 1.4 



28 



