oint of
similarity as regards their chemical conduct. On the other hand, oxygen
and sulphur, while quite different physically, have much in common in
their chemical properties.

About eighty elements are now known. If all of these should have
properties as diverse as do oxygen, hydrogen, and nitrogen, the study of
chemistry would plainly be a very difficult and complicated one. If,
however, the elements can be classified in groups, the members of which
have very similar properties, the study will be very much simplified.

~Earlier classification of the elements.~ Even at an early period efforts
were made to discover some natural principle in accordance with which
the elements could be classified. Two of these classifications may be
mentioned here.

1. _Classification into metals and non-metals._ The classification into
metals and non-metals most naturally suggested itself. This grouping was
based largely on physical properties, the metals being heavy, lustrous,
malleable, ductile, and good conductors of heat and electricity.
Elements possessing these properties are usually base-forming in
character, and the ability to form bases came to be regarded as a
characteristic property of the metals. The non-metals possessed
physical properties which were the reverse of those of the metals, and
were acid-forming in character.

Not much was gained by this classification, and it was very imperfect.
Some metals, such as potassium, are very light; some non-metals, such as
iodine, have a high luster; some elements can form either an acid or a
base.

2. _Classification into triad families._ In 1825 Doebereiner observed
that an interesting relation exists between the atomic weights of
chemically similar elements. To illustrate, lithium, sodium, and
potassium resemble each other very closely, and the atomic weight of
sodium is almost exactly an arithmetical mean between those of the other
two: (7.03 + 39.15)/2 = 23.09. In many chemical and physical properties
sodium is midway between the other two.

A number of triad families were found, but among eighty elements, whose
atomic weights range all the way from 1 to 240, such agreements might be
mere chance. Moreover many elements did not appear to belong to such
families.

~Periodic division.~ In 1869 the Russian chemist Mendeleeff devised an
arrangement of the elements based on their atomic weights, which has
proved to be of great service in the comparative study of the elements.