726 



BELL SYSTEM TECHNICAL JOURNAL 



To better visualize the fluctuations in p, the data of Table 1 are 

 shown graphically in Fig. 2a and Fig. 2b. It may appear that during 

 the months of May and October there existed some assignable cause of 

 variation in the production process of Type A apparatus. The same 

 may appear to be true for Type B apparatus during the months of 

 February and November. 



MONTHS 

 Fig. ia. Apparatus Type A 



UPPER LIMIT 



MONTHS 

 Fig. 3i. Apparatus Type B 



We shall see upon investigation that there is evidence of lack of 

 control of apparatus Type A but not any evidence of lack of control of 

 apparatus Type B. 



