Names, then, for numerals in the American languages differ as much as the natural objects from which they may be derived, the separation from the parent-stock of the tongues in which they occur having taken place before the evolution of fixed absolute and abstract terms. _The verb-substantive._--In the Indo-European languages the verb-substantive agrees even where other words differ; the English _be_ is the Latin _fu-_; the German _ist_ is the Greek ἐστ-ι; the English _am_ is the Latin _sum_, and the Greek εἰμι. This induces us, in languages where there is no such agreement, to argue in favour of a fundamental dissimilarity. And naturally. Tongues as far apart as the English and Sanskrit _agree_, where tongues as close to each other as the Adahi and Chetimacha _differ_. But to expect likeness on this point simply because we find it in Europe and Asia, is to make bricks without straw. In most of the American languages, an idea so abstract as that conveyed by the verb-substantive has yet to be evolved; in other words, there is no verb-substantive at all in the generality of them: according to some writers, it is wanting in all. Such are some of the facts and suggestions which help to account for the glossarial difference between the American languages, a phænomenon which, even though occasionally overstated, is still a reality to a certain degree. I am fully aware that, at the first view, they seem to prove too much; _i.e._ they seem, by accounting for the differences, to admit them; just as, in common life, the person who excuses himself for an imputed action, admits the truth of the imputation. How far this is the true view will be seen after the notice of some of the antagonistic phænomena of agreement in the way of grammatical structure. _Negative points of agreement._--_Case-endings_, properly so called, are either rare or wanting throughout the American tongues. Possession is expressed by the pronouns; just as if we said, _father his_, or _pater suus_ instead of _patri-s, ather-'s_. In like manner the pronoun expresses the objective relation; _I strike him horse_=_ferio equu-m_. _Signs of number_, properly so called, are wanting. The general American equivalent for such a form as the _-s_ in _patre-s_, or _father-s_, is a word signifying _number_, as _father many_=_father-s_. _Signs of gender_, properly so called, are wanting. This, however, is no more than what occurs in the English adjective. _Signs of the degrees of comparison_ are wanting. This, however, is no more than what occurs in the French adjective. Notwithstanding, however, this list of negations--a list capable of being considerably increased--the American grammar is complex; a fact which brings us to the positive characteristics of the language in question. These, also, are very general.