The assignment of a void patent presents a particularly interesting case of the non-existence of the thing concerning which the parties contract. Such an assignment purports to transfer the exclusive privilege or monopoly, under grant of the state, to make, use, and vend, and to authorize others to make, use, and vend, the subject matter of an invention. If for any reason the grant of the patent is void, the exclusive privilege or monopoly does not exist. And since there is no implied warrant of the validity of the grant,2 it ought to be held that the assignment is void for non-existence of the subject matter, and that money paid in misreliance upon its validity may be recovered. Such, apparently, is the law in the United States.3 But in England and Canada it is held that the assignment of a patent is valid and enforceable though the patent itself is void.4 This view probably rests upon the theory that the assignment purports to transfer, not a monopoly, but merely the assignor's rights under his letters patent, whatever they may turn out to be, and that the assignee therefore assumes the risk of the invalidity of the grant.1 If the assignment is so worded as to indicate that the parties regard the patent as of doubtful validity, such a construction may be justified.2 In the absence of such evidence it seems somewhat strained.

1 See Keener, "Quasi-Con tracts," p. 118. And compare Stein-back v. Rhinelander, 1803, 3 Johns. Cas. (N. Y.) 269; New Holland Turnpike Co. v. Ins. Co., 1891, 144 Pa. St. 541; 22 Atl. 923.

2 Hall v. Conder, 1857, 2 C. B. N. S. 22; Hiatt v. Twomey, 1836, 1 Dev. & Bat. Eq. (21 N. C.) 315; 3 Robinson, "Patents," Sec. 1232.

3 In Darst v. Brockway, 1842, 11 Ohio 462, it is held that money paid by the assignee may be recovered. Hiatt v. Twomey, 1836, 1 Dev. & Bat. Eq. (21 N. C.) 315, contra. In the following cases it is held that the invalidity of the patent is a defense to an action by the assignor to recover the purchase price or to an action on a note for the purchase price: Dickinson v. Hall, 1833, 14 Pick. (Mass.) 217; 25 Am. Dec. 390; Lester v. Palmer, 1862, 4 Allen (Mass.) 145; Harlow v. Putnam, 1878, 124 Mass. 553; Keith v. Hobbs, 1878, 69 Mo. 84; Herzog v. Heyman, 1897, 151 N. Y. 587; 45 N. E. 1127; Geiger v. Cook, 1842, 3 Watts & S. (Pa.) 266. And see Hamilton v. Park, 1900, 125 Mich. 72; 83 N. W. 1018.

4 Hall v. Conder, 1857, 2 C. B. N. S. 22; Liardet v. Hammond, etc., Co., 1883, 31 W. R. 710; Vermilyea v. Canniff, 1886, 12 Ont. 164.

A license, in the law of patents, is not, accurately speaking, a transfer of the licensor's monopoly, but a permit to make, use, or vend the thing patented under protection of the patent. Essentially it is merely an agreement by the licensor that he will not hinder or attempt to prevent the licensee from making, using, or vending the thing upon which the licensor claims a monopoly.3 The validity of the patent may be an assumed fact, but it is not a fact the non-existence of which makes the contract impossible of performance, since all that the contract requires is non-interference by the licensor. Consequently, though the patent is void, the license is valid and enforceable so long as the licensee continues without molestation to operate under it.4

1 In Hall v. Conder, supra, Williams, J., said (p. 42): "They contracted for the patent such as it was, each acting on his own judgment," and distinguished the earlier case of Chanters v. Leese, 1838, 4 Mees. & W. 295, upon the ground that there the assignor expressly contracted that the assignee should have the exclusive right to sell certain things for which the patents had been obtained.

2 See Johnson v. Willimantic Linen Co., 1866, 33 Conn. 436, in which the court said ( p. 443): "Where, as in this case, there is . . . clear evidence on the face of the instrument that the parties mutually contemplated the possibility, if not the probability that the patent was invalid, and provided by the form of the instrument and its stipulation for the contingency, it is very clear that there is no ground on which the vendee can be permitted to set up a failure of consideration." See also, Gilmore v. Aiken, 1875, 118 Mass. 94; Herzog v. Heyman, 1897, 151 N. Y. 587 ; 45 N. E. 1127.

3 See Heaton, etc., Co. v. Eureka, etc., Co., 1896, 77 Fed. 288, 290; 25 C. C. A. 267; 47 U. S. App. 146; 35 L. R. A. 728.

4 Lawes v. Purser, 1856, 6 ELL. & B. 930; Crossley v. Dixon, 1863, 10 H. L. Cas. 293; Holmes v. McGill, 1901, 108 Fed. 238; 47 C. C. A. 296; Rhodes v. Ashurst, 1898, 176 111. 351; 52 N. E. 118; Jones p. Burnham, 1877, 67 Me. 93; 24 Am. Rep. 10; Standard Button, etc., Co. v. Ellis, 1893, 159 Mass. 448 ; 34 N. E. 682; Strong v. Carver Cotton Gin Co., 1907, 197 Mass. 53; 83 N. E. 328, 330; 14 L. R. A. (N. S.) 274; Marston v. Swett, 1876, 66 N. Y. 206; 23 Am. Rep. 43. See also Ross v. Dowden Mfg. Co., 1909, 147 la. 180; 123 N. W. 182.

In many cases allowing the recovery of royalties under void patents, the decision is rested upon the doctrine of estoppel. Unless the licensee has acknowledged the validity of the patent (as in Hyatt v. Ingalls, 1891, 124 N. Y. 93; 26 N. E. 285), the doctrine would seem to be inapplicable.

It follows (upon the principle of Sec. 18, ante) that royalties paid by the licensee may not be recovered.1

The distinction between an assignment of a patent and a license has frequently been overlooked. The leading English case of Taylor v. Hare,2 which held that royalties paid by a licensee under a void patent could not be recovered, is criticised by Professor Keener,3 apparently upon the theory that the licensee had bought "the right to manufacture under the letters patent," which since the patent was void, the licensor did not have to sell. If this is the true theory of the nature of a license, the criticism is deserved; but as has been said, the courts have construed licenses to be merely agreements by the licensor not to interfere with the licensee.