10-gram weights, or the two 100-milligram weights) are actually equal
and interchangeable, or that the 500-milligram weight is equal to
the sum of the 200, 100, 100, 50, 20, 20 and 10-milligram weights
combined, and so on. If discrepancies of more than a few tenths of a
milligram (depending upon the total weight involved) are found, the
weights should be returned for correction. The rider should also be
compared with a 5 or 10-milligram weight.

In an instructional laboratory appreciable errors should be reported
to the instructor in charge for his consideration.

When the highest accuracy is desired, the weights may be calibrated
and corrections applied. A calibration procedure is described in a
paper by T.W. Richards, !J. Am. Chem. Soc.!, 22, 144, and in many
large text-books.

Weights are inevitably subject to corrosion if not properly protected
at all times, and are liable to damage unless handled with great care.
It is obvious that anything which alters the weight of a single piece
in an analytical set will introduce an error in every weighing made
in which that piece is used. This source of error is often extremely
obscure and difficult to detect. The only safeguard against such
errors is to be found in scrupulous care in handling and protection
on the part of the analyst, and an equal insistence that if several
analysts use the same set of weights, each shall realize his
responsibility for the work of others as well as his own.


BURETTES