478 HARVARD LAW REVIEW. nect the whole with the defendant's signature? This is a question in the law of incorporation by reference, not in the law of evidence. The views expressed by the English courts seem to resolve themselves into one of the two following propositions: (i) that the signed document must contain a reference to some other writing, or (2) that it need only refer to a transaction, of which the other writing in fact forms a part. At the beginning of this century the first of these views was modified by a requirement that the identity of the writing referred to must appear from the reference itself In Boy dell v. DncmmoJid, 11 East, 142 (1809), there was a prospectus of sale, and the defendant signed a book headed " Shakespeare Subscribers, their Signatures." During the argument Lord EUenborough said, ** What is there in the title to refer to the particular prospectus rather than any other ? If it had referred to the partic- ular prospectus then published, it would have helped the plaintiff over his difficulty." This early modification, however, was dispensed with in Feirce v. Corf, 9 Q. B. 210 (1874), where the court adopted without limi- tation the view expressed in the first of the above propositions : " On the document itself there must be some reference from one to the other, leaving nothing to be suppHed by parol evidence except the identity, as it were, of the document." Though couched in terms of evidence, this amounts to saying tiiat when the signed paper contains a reference to some other writing, whether to a particular one or not, and there exists another writing which is shown (no matter how) to be the one referred to, the secoiid writing can be incorporated into the first, and the defend- ant's signature will thus be annexed to both in a way satisfactory to the statute. After Peine v. Corf there came a series of cases, in which the courts, though professing to follow the first rule, seem in substance to have adopted the second, namely, that if a signed document refers to a trans- action generally, it will incorporate any document connected with that transaction. In Lo?ig v. Millar, 4 C. P. Div. 450 (1879), the signed document referred to "the purchase," — not to any document connected with the purchase, but to the transaction generally. The court there incorporated a pa- per containing the plaintiff's agreement to buy, and tried to reconcile the case with the former rule by construing " the purchase " to mean "the agreement to purchase." In Studds v. Watson, 28 Ch. Div. 305 (1884), the second proposition was squarely stated by North, J , as follows : "There is a parol agreement proved to which both these documents refer. All the terms of this parol agreement are found in one or the other of the two documents, and are in themselves sufficient to constitute a good memorandum within the statute. ... I do not think it is necessary in this case that these two documents should refer the one to the other." In that case, to be sure, both papers were signed, but the court appears not to have noticed the fact. Again, in Oliver v, Himting, 44 Ch. Div. 205 (1890), there was a letter acknowledging the receipt of a check "on account of the purchase money for the Fletton Manor House estate," and this was held to refer to and incorporate a previous memorandum relating to this transaction. There was here no reference to a document, but merely to the transaction of which the document was a part. Thus at this period the English courts had abandoned the first rule and had virtually adopted the second. In 1895 the question came up again in Potter v. Peters, supra.