plainant's patent, but, on the contrary, that changes must be made in them before it is even pretended that they will show this invention. The supreme court, in Topliff v. Topliff, 145 U. S. 156, at page 161, 12 Sup. Ct. 828, has said:

"It is not sufficient, to constitute an anticipation, that the device relied upon might, by modification, be made to accomplish the function performed by the patent in question, if it were not designed by its maker, nor adapted, nor actually used, for the performance of such functions."

The force of this ruling, and the similar ruling in Clough v. Barker, 106 U. S. 175, 1 Sup. Ct. 188, is made manifest, in its practical application to the rights of parties, by the reflection that all earlier patents set up in defense against a later patent sued upon are but the record evidence of the status the art has reached. The rights under such later patent are subject to what this record evidence actually shows. Το change this record, by permitting theoretical modifications of these earlier patents, would be the same, in principle, as to change, by interpolation or modification, any other evidence between the parties.

The witness Minnis (answer 24) discusses at great length the question of whether there is anything novel and of a patentable character in the patent in suit, in view of patents Nos. 292,865, 292,866, 330,059, 330,060, 294,482, and 269,238. He does not say that he finds the invention of the patent in suit in either, or all, of these patents, but would make it appear that, by changing some or all of them, or by combining some of them together, thus resorting to the process of modifications and reorganizations, he would then arrive at this invention. For in

stance, he says:

"Thus, it will be seen, by means of these four patents, we have been brought up step by step to an arrangement and combination almost identical with that shown in the patent in suit."

But this "almost," with its indefinite bounds, forms the missing link, and, it is thought, brings this form of attack on complainant's patent substantially within the rule pronounced in Topliff v. Topliff, supra, and Clough v. Barker, supra, and is also within the objection to such a course pointed out by Judge Coxe in the case of Johnson v. Railroad Co., 33 Fed. 500, where he says:

"The defendant assembles every similar device, description, or suggestion, in the particular art not only, but also in analogous, and even remote, arts. Everything which has the least bearing upon the subject is brought in and arranged by a skillful expert in an order of evolution which resembles most clearly the invention which is the subject-matter of attack. Having thus reached a point where but a single step, perhaps, is necessary to success, and knowing from the inventor exactly what that step is, the expert is asked if the patent discloses invention, and, honestly no doubt, answers in the negative. There is always the danger, unless care is taken to divest the mind of the idea added to the art by the inventor, that the invention will be viewed and condemned in the light of ascertained facts. With his description for a guide, it is an easy task to trace the steps from the aggregation to the invention."

"In the law of patents, it is the last step that wins," says Justice Brown in The Barbed Wire Patent, 143 U. S. 275, 12 Sup. Ct. 443, 450. Bearing in mind that defendants' witness Clark distinguishes between the patent in suit and such of these alleged prior patents as he considers, and that defendants' other witnesses, Paca and Minnis, do not point out the invention in suit in these earlier patents, it is pertinent to consider the rebuttal offered by complainant, through Scribner, the patentee, who discusses all of these alleged earlier patents, as also patent No. 306,669, offered by defendants in connection with the Clark deposition, but not discussed by any of defendants' witnesses. Scribner, in answers 1 and 7, points out that these patents, three of which are his own, do not contain the invention of the patent sued on. Complainant's witness Haskins also considers these patents, and testifies

that he does not find in them the invention in question; and I concur in this view.

This discussion disposes of the third proposition of defendants, to the effect that the patent in suit should not have been granted, because the same subject-matter was disclosed in prior patents to the same patentee, and that his rights were relegated to his remedy by reissue, and not to an original patent. It is not clear, from my view of the "prior patents" to this patentee, how they could be reissued for the subject-matter covered by this one patent.

These considerations dispose, also, of the fifth contention of defendants that the patent in suit is void, as to the claims in issue, under the decision in Miller v. Manufacturing Co., 151 U. S. 186, 14 Sup. Ct. 310, and Thomson-Houston Electric Co. v. Winchester Ave. Ry. Co., 71 Fed. 192. In Miller v. Manufacturing Co., the later patent contained the same mechanism as the earlier patent to the same patentee, and the essential or only difference in the two patents was that the earlier claimed the mechanism in respect of its two functions, and the later claimed the mechanism in respect of one of those functions; the case presenting one in which there was no division of mechanism, but only a division of functions as between the two patents. There is nothing of that sort as between the patent in suit and any of the patents offered by the defendants.

In the additional brief filed on behalf of defendants, considerable space is devoted to showing that patent No. 321,390, dated June 30, 1885, issued upon an application filed October 25, 1879, contains a substantial disclosure of the separable contact, cutting-out feature; defendants contending that everything disclosed by the patent in suit, except, possibly, the peculiar form of the apparatus and the particular form of the test circuit, belongs to the patentee's earlier patent, No. 321,390, and say that the idea of excluding one set of contacts at the same time connection is made with another set is clearly set forth in this patent, No. 321,390, and that the patent in suit was therefore void, or anticipated. This patent, No. 321,390, describes its spring jacks

thus:

"One part, B', of the switch, I term the 'line' portion, and the other part, B, 1 term the 'test' portion."

These portions are not used simultaneously, but each is used, when occasion requires, for its own particular purpose.

Defendants' witness Minnis (answer 23), speaking of these contacts in patent No. 321,390, says that:

"Both of these contact switches being of a separable character, the insertion of a plug would still leave in the circuit those contact switches between the plug and the point of entry of the line into the switch board."

It appears, then, that upon entering the plug of a connecting cord into any of the series of switches there would still be left, in the talking circuit, the separable contact points of all the intervening switches located between the plugged switch and the point of entry of the talking line into the switch board. This statement accords with the testi

85 F.-42

mony of the patentee, Scribner, in his answer to question 7, wherein he

says:

"In this patent, No. 321,390, instead of a single switch upon each multiple switch board, there are two switches for each telephone line, and these two switches are insulated from each other, and the contact points of all the switches of a given telephone line are included in the line circuit of the telephone line. A connection of any two lines of this system gives a talking circuit through these contact points, thus introducing an objectionable resistance,-a defect which the patent in suit was designed to and does obviate. Neither the construction nor the result of the patent in suit can be possibly secured by any employment of the apparatus of my patent No. 321,390."

The witness Haskins, for the complainant, in answering question 8 in his deposition, says:

"I will next consider Mr. Scribner's patent, No. 321,390, of June 30, 1885. This patent also contains the same objectionable feature of including in the talking circuit the contacts of the switches when two subscribers are connected for conversation, and accordingly has no significance."

Then Minnis, in his twenty-third answer, speaks of certain changes to be made in patent No. 321,390, to make it accomplish the result effected by the patent in suit. For instance, he says that:

"A substitution of one bolt with two sets of contact switches would be an absolute equivalent of the two bolts, each having a contact switch."

Again, he says:

"Going a step further, and substituting for this bolt with two sets of contact switches [of] a bolt with one set of contact switches and one solid switch contact, and we have an arrangement under which the insertion of the plug in any switch would leave no separable contacts included in the circuit thus established."

If patent No. 321,390 covers broadly the identical invention covered by the patent in suit, as is claimed by defendants' counsel, why the substitution or change of bolts and contact switches, as suggested by witness Minnis? Why any change of construction in the arrange ment shown in patent No. 321,390, if it possesses all the functions of the patent in suit? "The grant of letters patent is prima facie evidence that the patentee is the first inventor of the device described in the letters patent and of its novelty. The burden of overcoming the prima facie case made out by the production of the patent is upon the defendant, and the defense of want of novelty must be clearly established before a court will be justified in setting aside the patent on this ground. Not only is the burden of proof to make good this defense on the party setting it up, but it has been held that 'every reasonable doubt should be resolved against him.'" Cantrell v. Wallick, 117 U. S. 689, 6 Sup. Ct. 970; Palmer v. Village of Corning, 156 U. S. 342, 15 Sup. Ct. 381.

This disposes of all of defendants' propositions, except the fourth and sixth, which go to the point that there is no infringement by defendants of the patent sued on, and to the further point that the claims can only be saved by a narrow construction, restricting them to the exact form, construction, 'and arrangement described by the patent in suit and illustrated therein. It does not seem to admit of doubt that this patent accomplishes a new result, namely, the direct connection

of two subscribers' lines through the usual plugs and cords, at the same time cutting out of the talking circuit thus formed all the separable contacts belonging to such lines, and cutting out the conductors leading to the individual annunciators. The instrumentalities employed to accomplish this result are not, individually, new instrumentalities, but they are brought together in a new relation, combination, or organization, and thereby produce this advance in the art of telephony, and which, according to the evidence in this case, is meritorious in respect to its practical utility, and the advantages it affords in the actual use of telephones, by simplifying the apparatus in telephone exchanges, and by improving the tone and efficiency of talking circuits, through the elimination of objectionable resistance to the low current employed in such circuits. As defendants' Exhibit A is accepted by both parties to the suit as a substantially correct representation of defendants' apparatus, as used in their exchange, that exhibit has been adopted by me as the basis of comparison with the patent sued on.

Defendants' witness Clark, in answer 23, says:

"The said defendants' Exhibit A represents correctly the apparatus as used in the switch board at Mobile, Alabama, by the Home Telephone Company. This is a correct drawing."

Clark was further asked:

"In defendants' system, the circuit of each telephone line may be traced normally from the subscriber's station, through the separable contact of the different spring-jack switches of the line, and thence through the line annunciator, and thence back to the subscriber's station. Is not this correct? A. Yes, sir (cross answer 78). Referring to Fig. 1, for example, of the patent in suit, is not each spring jack of each line provided with three contacts, one contact being normally open, and the other contacts being normally touching each other or closed together? A. Yes, sir (cross answer 79). Is not the same true as to the spring jacks of defendants' system? A. Yes (cross answer 80). When a connecting plug is inserted in a spring jack of the patent in suit, is not the normally open contact of the spring-jack switch closed to the plug to form a circuit to the telephone line, while at the same time the separable contacts are disconnected, so as to cut off the portion of the line or the branch which contains the line annunciator? A. Yes, sir (cross answer 81). Is not the same true of defendants' system? A. Yes, sir (cross answer 82). In the case of the patent, and in the case of defendants' system, when two lines are connected together, none of the separable contacts of such lines are included in the united circuit of the two lines; that is, in the talking circuit. Is this correct? A. Yes, sir (cross answer 83)."

And said witness Clark was further asked:

"Referring to Fig. 1 of the patent in suit, do you not find a branch from the telephone line connected with the normally open contact of the spring-jack switches of such line? A. Yes, sir (cross answer 87). Referring to defendants' apparatus, as illustrated in defendants' Exhibit A, are not their contact springs. el and e3, of the spring-jack switches of the telephone line of station A, and are not these contacts, el and e3, both branched from the said telephone wire, and are they not normally open? A. Yes, sir (cross answer 88)."

Paca, another of defendants' witnesses, stated that:

"Exhibits A and B, diagrams of defendants' apparatus, correctly represent the arrangement of the circuits and apparatus at the exchange of the Home Telephone Company in Mobile."