512 HARVARD LAW REVIEW. remembered that the classification is based upon a classification of intentions, and is therefore not an organic division, such as that into genera and species, but is rather a mechanical division, such as would be that of books classified according to the number of their pages. It is of course immaterial what vocable we use to mark these differing relations. It is material, however, that, when once we have made a choice of words, we should thereafter confine the word chosen within the limits of accurate definition. Cases, for example, where there is a transfer of title and a preservation as an entirety of a fund should not be confounded with cases where these play no part. Let me now, therefore, for the sake of accu- rate discrimination, call attention in more detail to the differences between the last two cases. In the first of the two, that is, the fourth of the supposed cases, the relation between A and myself was dependent solely upon the communications which passed between us two. A became my servant just to the extent to which I gave him directions. In the second of these, that is, in the fifth of the supposed cases, we find that A's agency is inchoate and ineffectual until B has learned of the communications wherein I gave A the right to act in my behalf. It may happen that B is informed of only a part of these communications, nor is it neces- sary that he should know more of them than will suflfice to satisfy him of A's representative character. So much, however, he should know, or else he will refuse to deal with A, and will insist upon direct communication with me, or upon none at all. It follows, therefore, that the intended relation of A to me is not complete in fact, or intelligible in thought, until there are, first, instructions communicated by me to A, and, second, a statement of those instructions made to B. This second element at once distinguishes by a wide gap, the fifth case from the fourth. The latter, being a case where only two persons are necessarily included, is an in- stance of a bilateral relation, while the former, where three are necessary, is an instance of a trilateral relation.^ 1 Professor Huffcut, in his book, the most recent treatise on agency, very nearly perceived the true nature of these relations. Thus he says : " The fundamental dis- tinction between an agent and a servant lies in the nature of the act which each is authorized to perform. An agent represents the principal in the performance of an act resulting in a contractual obligation, or an obligation springing from contract relations. A servant represents the master in the performance of an act not resulting in contractual obligation." Huffcut on Agency, § 4 (1895). In his subsequent illustrations in the same section, he distinctly shows that the
