are aggregated, rather than integrated; the root or theme is held apart from the affixes, and these from one another, with a distinct apprehension of their separate individuality. As Professor Müller well expresses it, while Indo-European language, in putting two roots together to compose a form, sinks the individuality of both, the Scythian sinks that of but one, the suffix. The process is not, in its first stages, diverse in the two families, since every Indo-European form began with being a mere collocation, and, in a large proportion of cases, the root maintains to the end its integrity of form and meaning: the difference is one of degree rather than of kind; of the extension and effect, rather than the essential nature, of a mode of formation: and yet, it is a palpable and an important difference, when we compare the general structure of two languages, one out of each family.
The simple possession in common of an agglutinative character, as thus defined, would certainly be a very insufficient indication of the common parentage of the Scythian tongues; mere absence of inflection would be a characteristic far too general and indeterminate to prove anything respecting them. They do, however, present some striking points of agreement in the style and manner of their agglutination, such as might supplement and powerfully aid the convincing force of a body of material correspondences which should be found wanting in desired fullness. The most important of these structural accordances are as follows.
In the Scythian languages, derivation by prefixes is unknown; the radical syllable always stands at the head of the word, followed by the formative elements. The root, too, to whatever extent it may receive the accretion of suffixes, itself remains pure and unchanged, neither fused with them, nor euphonically affected by them: throughout the whole body of its derivatives, it has one unvarying and easily recognized form. It would appear, however, on theoretical grounds, that this fundamental characteristic, of the inviolability of the Scythian roots, must be admitted with some grains of allowance: since, if root be kept absolutely separate from ending, and changeless, we should, on the one hand, look for a much closer coincidence of roots than we
