


WEIGHT VARIATION OF PACKAGE FOODS. 11 
of comparison, the practical limit of each of these distributions is 
considered to be the corresponding figure in the next to the last 
column of Table 5 and of Table 6, ‘Calculated maximum error on a 
single package.” 
TABLE 7.—Observed variations of good commercial practice. 








a eoee 
Average eviation 
. Declared . Average F of single 
Experiment No. Product. weight. Aim.! founds ae packages 
from 
average. 
Ounces. | Ounces. | Ounces. | Ounces. | Ounces. 
STS ue) Ae i le ee oe Cayenne pepper-..-. 0. 75 0. 78 0.76 | —0.02 0. 050 
BO DID Soe ae arse arash sicvsis Hie ee UGE S Soneee Beep ae sae 1.33 1.40 1. 40 -00 - 031 
CU Ca site BOE Se eee ees eee = do Sceavbososeou6 1.33 1.35 1.37 02 - 042 
SIDS re eis ore Pe cicge acta ce Se aiain'sre Gin Sei asso ease 2.00 2. 03 2.05 02 - 025 
SOUG eee ei- sce eens see coe Greekjsagesnecs see 2.00 2.02 2. 04 . 02 - 024 
SO0f a sie. {ssc en = se ese es Ground mustard.... 2.00 2.05 2.04} — .01 015 
SUE. eSocosscesoseessoepacopeec Ginger eoccssoac-ne- 2.00 2.04 2.03 — .01 - 025 
Averages of approximately 2-ounce packagés. .............----------.---------- - 003 . 030 
AV CEAPOeVIA ONUROMN average mnINUS AlNIs = \seces sone Dees nee oo non dee aceee eens 014 
SOUOSae ae soe es sce oen tec e. Pickling spice....... 3. 50 3.55 3.56 01 -031 
Sosa ae Seeian he ot sce eeewaraiee Cinnamon........... 4.00 4.05 4.12 07 -018 
4 Bee arses ciate aa Staats we Ginger sess sos oe 4.00 4.07 4.09 02 - 026 
sc] Oy is i Bee a eae ene S Cinnamon’ o5.5-e5 4.00 4.05 4.08 03 017 
Averages of approximately 4-ounce packages. ...........22-000see-- ee eee eee eee - 032 - 023 
AVEFASe COVIAtON ONO AVELALE MINUS AN: «cos ace sese ceo nsec snaccens das sametaesees = - 032 
SSS Qieepaoaeee eae ae eee Dessert powder ..... 7.00 7.05 7.08 03 - 027 
Sa 6 BacsuaGeee senses ee eee eaemee doh aes 8s 7.00 7.03 7. 20 2.17 . 061 
Gall Grae eee meee ee eee se OOe wast cesece ss 7.00 Ce Ee = eal 02 - 052 
2259) Dp a CRE e Rae eee Gound nutmeg..... 8.00 10. 70 10. 72 -02 . 027 
Averages of approximately 8-ounce packages. .......-.....------------.+2+--+-- - 023 . 042 
FAVeraco devintlonirom OlaVverage MINUS ALM. 5. s.5-2< sae o oe = saone cc ucasgcuccs cemeees . 023 
CO. on SSSR aE e See eee ee Pulverized coffee. -.. 1lb. 16. 07 16. 07 -00 . 024 
OSA eee oe eine sete eee es See RIG EROS Seen 1lb. 16. 02 16.10 08 031 
SUCH ec ese rea eS UI COP OMI ee 1 1b. 16.04 16.07 03 064 
Roe aera e cesBoee ese sess k Coleen cae sane 1Ib. 16.00 16.02 02 036 
Averages of 1-pound packages .......-.----.---- a Si ee ice Parnes - 032 - 039 
Average deviation from 0 average minus aim .............-.---.--..-------20- Boscoceecte - 032 
She sees seeecs sc. sSae-5s Colter sex conc scck | 3 Ibs 59. 16 | 59. 28 | 12 094 
AMEFAPOGG MAHON LOM: AVEFAse MINUS AlN... cas c~ series secs 5 oec case se cesenicemen deer - 120 
Uy ie Oe SE ee ee Ground cloves......- 5 Ibs. 79. 55 79. 55 -00 . 120 
A. es a a Ae ee Baking powder.....-. 5 Ibs. 89. 46 89.45} — .O1 218 
AV eTAMESEOMO-DOUNG. PAackAGeSs <0 asec cs scons cedinc wince sees eccuss i AE As — .005 . 169 
AVCTALOCS VIAhION ILO: Ola VErAseuMINUS ‘ALM. Sac cisnn cieeiseeleisid = cialis oe Sn ee eis clase sieeinee - 005 
SU Yay AS 58 ie ee ae ee a | Reppebe ses ua eee | 50 lbs. | 800.00 | 799. 30 | — 540 400 
AVEeravOGeviation irom 0 average MINUS aiMs Mss. sce cesid- soc ose e neces enase ces 70 
1 The operators were giving a slight overweight, the amount of which was estimated by observing the 
method of weighing and the sensibility reciprocal ofthe scale. Addition of this amount to the reading of 
the scale gave the aim. 
2 This operator persisted in giving excessive down weight, contrary to instructions. His results are 
not included in the average. 
3 This set was weighed with exceptional care by the packer, but all other conditions were commercial. 
COMPARISON OF CALCULATED RESULTS WITH OBSERVED RESULTS. 
Although the theory of probability provides for errors of infinite 
magnitude from the average of a normal distribution, in practice 
their magnitude is limited. The limitations may be measured by 
the calculated chances of occurrence of the maximum error. A 
