Roman state, of whatever race he might be; while from here it not only passed to the emperor (kaiser) of Germany, whose throne pretends to be the modern representative of that of Rome, but also to the autocrat (czar) of distant and barbarous Russia—thus becoming the equivalent of 'emperor' in two of the most important languages of modern Europe.
These examples are of themselves sufficient to place before our eyes the most important features in the history of significant change of words, the principal processes by which—even apart from combination or phonetic change, but yet more effectively in connection with these—the existing vocabulary of a language is adaptable to the growing knowledge and varying needs of those who use it. We see that, in finding a name by which to designate a new conception, we may either pitch upon some one of the latter's attributes, inherent or accidental, and denominate it from that, limiting and specializing for its use an attributive term of a more general meaning; or, on the other hand, we may connect it by a tie of correspondence or analogy with some other conception already named, and extend so as to include it the sphere of application of the other's designation; while, in either case, we may improve or modify to any extent our apprehension of the object conceived of, both stripping it of qualities with which we had once invested it and attributing to it others, and may thus pave the way to the establishment of new relations between it and other objects, which shall become fruitful of further changes in our nomenclature. These two, in fact—the restriction and specialization of general terms, and the extension and generalization of special terms—are the two grand divisions under which may be arranged all the infinite varieties of the process of names-giving. Some of these varieties and their effects, however, it will be desirable for us to examine and illustrate more fully, before going on to consider farther the general character of the process. We will not attempt in our illustrations a strictly systematic method, but will take something of the same freedom which linguistic usage assumes in dealing with the material of speech.
It is obvious how vastly the resources of a language for
