152 MAINE AGRICULTURAL EXPERIMENT STATION. 



with mercury as is also the burette. Then the mercury is 

 allowed to run slowly from the burette into the bottle until the 

 upper line of the mercury stands at the 5 per cent, mark on the 

 burette. The mark at which the upper line of the mercury 

 stands in the neck of the bottle is also noted and if it coincides 

 with the 5 per cent, mark, the graduation is correct. Another 

 portion is now run in until the 10 per cent, mark on the burette 

 is reached and so on until the 25 per cent, mark is reached. 

 Slight errors like one or two-tenths of 1 per cent, have been 

 passed unnoticed, but when there was an error of three-tenths 

 or more, the bottles have been thrown out, not because three- 

 tenths in practice is a serious error, but because it is better to 

 insist that manufacturers shall furnish goods up to their guar- 

 antee. 



For testing pipettes, an accurately graduated standard pipette 

 is used. It is filled with mercury and then the contents are 

 emptied into the pipette to be tested. If the point reached by 

 the mercury coincides with the mark on the standard pipette, 

 it is of course correct. All bottles and pipettes tested and found 

 correct have been marked with the letters O. K., while those 

 that are not correct are marked "off.'' 



Only about one-half of the creameries of the State have sent 

 their glassware to us, which indicates one of three things: (1) 

 That those who have not responded are not using the test; (2) 

 that they have obtained tested glassware from dealers, or 

 (3) they have not complied with the law. In all, 1,498 cream 

 bottles, 210 milk bottles and 96 pipettes were received prior to 

 January 1st. Thirty-nine of the cream bottles were found to be 

 more inaccurate than the prescribed limit, three-tenths of one 

 per cent, and were consequently thrown out. Twenty-four of 

 the number, however, were found in two small lots which evi- 

 dently came from some unreliable manufacturer, as the error 

 found was greater in those, in some cases over one-half of one 

 per cent., than in any other lots. Of the 210 milk bottles all 

 were correct, excepting 33 of one lot of 60. The errors in these 

 bottles varied from three-tenths to one per cent. The source 

 of these bottles could not be learned as they were found at the 

 creamery when the parties now occupying it took possession. 

 They were evidently made by some unreliable firm. 



