Since scientific knowledge is carried from one generation to another largely by means of written records, the terminology employed should be as nearly fixed as possible. Yet, as new theories are constructed and others modified or done away with, and as knowledge is increased, many of the terms that survive suffer important changes of meaning. Some of the stability of a fixed terminology, we may judge, is illusory, yet it has often been thought better that the same term be used in a different sense than that an entirely new term be adopted.