Page 2 - APPLIED INORGANIC ANALYSIS

PREFACE

During the past fifty years great advances have been made in the
development of methods for the determination of the elements and in
the clarification of the theoretical considerations on which analytical
procedures are based. As a consequence very accurate determinations
of many of the elements can now be made, provided they occur alone
or in very simple combinations. It must be confessed, however, that
corresponding advances have not been recorded in the reliability and
accuracy of analyses of the more or less complex mixtures in which
the elements normally occur. In other words, methods for separating
the elements remain much as they were, and the newer methods for
determining them do not differ greatly from the older methods as
regards their ''selectivity." Correct conditions for the determination
of aluminum by precipitating with ammonium hydroxide have been
established, but the analyst still has the problem of first separating
the aluminum from the numerous elements that usually accompany it
and are also precipitated by ammonium hydroxide.
After all is said and done a good analyst is primarily a chemist
who can quantitatively manufacture pure chemicals. The most elabo-
rate precautions in weighing and measuring and in the final determina-
tion of the element go for naught if the analyst has failed to prepare
quantitatively the compound on which the calculation is based, and to
free it quantitatively from all other compounds that would affect the
calculation. That this may be an extremely difficult task can be judged
by the errors that have been made in determinations of the atomic
weights, in spite of extraordinary expenditures of time and effort. For
example, in 1921 the accepted atomic weights of aluminum, silicon and
antimony were 27.1, 28.3, and 120.2, respectively, In 1925 these were
changed to 26.97, 28.06 and 121.77, the differences in parts per thou-
sand being 4.8, 8.6 and 12.9. Today, the atomie weight of zirconium
is rounded off to 91 because no one knows how much hafnium was
contained in the zirconium compounds on which the determinations of
the atomic weight of zirconium were based.
In addition to the ordinary difficulties, mineral analysis is very
often complicated by the faet that ideally pure minerals are rare. Not
only does isomorphism often introduce components of related members
of an isomorphous group, but solid solutions of unrelated minerals may

V