16 
BULLETIN 1009, U. S. DEPARTMENT OF AGRICULTURE. 
sets were filled by hand and some by machine. The maximum, mini- 
mum, and average volumes show the range of volume. As in 'Table 
3, the average of the average deviations of various sizes is an index of 
the variation in volume of food in the bottles of the size mentioned, 
and the figures in this column are used to compute the chance of 
occurrence of the calculated maximum variations in the volume of 
bottled foods. By computations similar to those used in Table 4, the 
chance of occurrence of variations in the volume of food in the bottle 
larger than the calculated maximum variations has been computed 
and compiled in Table 6. 
Table 
-Chances of occurrence of variations in the volume of food in the 
bottles larger than the calculated maximum -variations. 
Calculated 
i 
maximum 
Average 
X 
Chance of 
variation, 
deviation 
Probabil- 
ity. 
occurrence 
Capacity. 
individual 
X0.8453. 
r 
based on 
bottles. 
observed 
data. 
"(I) 
(r) 
(/) 
Fl. oz. 
Fl.oz. 
Fl.oz. 
lin— 
0.75 
0.10 
0.029 
3.45 
0. 9800 
50 
6 
.24 
' .227 
1.06 
.5254 
2 
8 
.29 
.272 
1.07 
.5295 
2 
15 
.39 
.131 ' 
2.98 
.9556 
23 
20-20 
.39 
. 236 
1.65 
.7342 
4 
63 
1.16 
.269 
4.31 
.9964 
278 
128 
1.94 
..686 
2.83 
. 9437 " 
18 
The results of the calculations in Table 6 are given in the last 
column. The average of the last column is 54, showing that, as a 
rule, one bottle in 54 will fall without the calculated maximum varia- 
tion. With the exception of the 63-ounce size, where the chance of 
occurrence of the calculated maximum is 1 in 278, the results indi- 
cate that there might be a frequent occurrence of variations larger 
than the calculated maximum variation. It should be noted, how- 
ever, that soda water is perhaps the most difficult food to bottle ac- 
curately, sinCe it is carbonated. The vinegar bottled in the estab- 
lishments in which these data were collected was bottled hot, thereby 
giving chances for additional variation due to cooling. When it is 
considered that the aim of the bottler- was changed in some cases 
from his normal height of fill, it is evident that a number of factors 
have been introduced to cause variation which would not be met in 
bottling uncarbonated liquids under conditions which have been 
customary for some time, as would be the case if the bottler were 
following the good commercial practice outlined in this bulletin. 
These data represent variations which would be met under very ex- 
treme and difficult conditions. Under more favorable circumstances 
the chances of occurrence as calculated in Table 6 would be re- 
duced. 
