RKUABI/JIV OF HOLDING TIME MEASUREMENTS 



.^71 



Finally there may be some question as to the busy season, its length 

 and stability. Plotting of the same data for the 9-12 a.m. hours as in 

 Fig. 4 A and 4B will help to decide these points. The morning hours' holding 

 time averages for each day of the week are plotted for several weeks during 

 the suspected busy season. Wide changes in the load through the passage 



155 

 150 

 145 



140 



en 



f 135 



O 



O 



UJ 130 



in 



Z 160 

 to 



UJ 



2 155 



< 140 



-V--^. 



3 4 5 



NUMBER OF WEEK 



Fig. 4 — Seasonal trend in holding time averages, 105 9-12 a.m. hours, Newark 



of weeks can be noted by eye. In the case illustrated there appears no 

 consistency of movement. Applying the identical test used in the previous 

 paragraph for the day-of-the-week changes, we find in Fig. 4C that in the 

 first five- week period no movements of significance took place. However, 

 the sixth week, which followed the first five after an interval of about 

 six weeks, showed a significant drop suggesting the approach of a lower 



