READING TEST BOTTLES. 305 



(a) of the fat column to the bottom of the meniscus 

 (b), or from (a) to half way between (b) and (d). 

 Do not add or subtract anything in practical work. 

 When the bottom of the fat column is smooth and 

 well defined, and the meniscus is clear and well de- 

 fined, half of the meniscus may be included in the 

 reading (b-d). That is, read from (a) to (d). "When 

 the bottom of the fat column is irregular and the 

 meniscus not well defined, read from (a) nearly to 

 (d), between (b) and (d). Since the graduations 

 are divided into whole per cents and the neck is so 

 wide, the bottle is not at all desirable for accurate 

 work. 



For convenience in illustrating, the Fig. V. shows 

 graduations in .5 per cent instead of 1 per cent, the 

 way the bottle is actually graduated. Since the 

 cream bottle in Fig. V. is only graduated into 1 per 

 cent, and 1 per cent on this bottle occupies only 

 about 1/16 of an inch, it is plain that a slight error 

 in reading this bottle would make a great difference 

 in the results. 



83. Comparing a 30 per cent, 6 inch, 18 gram, 

 cream test bottle with bottles shown in Figs. II., 

 III., IV., V. The 30 per cent, 6 inch, 18 gram bot- 

 tle, Fig. VI., is graduated in % per cent and 12 per 

 cent occupies a space of 1 inch. Therefore, .5 per 

 cent occupies 1/24 inch, which is about the same as 

 bottle in Fig. IV. This is a very good bottle, unless 

 cream tests more than 30 per cent when an 18 gram 

 sample must be used. 



When the scales are sensitive and the sample well 

 mixed and correctly weighed, it may be advisable to 



