O'Brien v. Railway Co., 102 Wis. 628, 78 N. W. 1084. There must be some direct evidence of a fact, or circumstances from which a jury would be warranted in saying that the inferences there from clearly preponderate in favor of the existence of the fact, in order to take the case to the jury. There must be sufficient evidence to remove the question from the realm of mere conjecture. Hyer v. Janesville, 101 Wis. 371, 376, 77 N. W. 729. When there are sufficient facts to go to a jury its finding is conculsive, unless reasonable men must have drawn from the facts a different conclusion. M. K. & T. Ry. Co. v. Byrne, 100 Fed. 359, 40 C. C. A. 402. A number of conjectures may be made from the evidence, as to the cause of the accident. The noise of an elevated train, or street car, of the boys playing or throwing stones, or some other noise or act, may have caused the fright, followed by such a pull as would have broken a good tie-strap. But these are all possibilities, of uncertain character. On the other hand, responsible facts appear-the weak strap, tied low, its being broken before the horse reared, the fright, and all reasonable inferences which the jury might draw from such facts, and others mentioned. These are enough, we think, to sustain the verdict. The request of defendant to charge that, if the jury found that the fright of the horse was so great that no strap, such as an ordinarily prudent man would have used, would have held the horse, then the verdict should be for defendants, does not tally with the fact that the fright appeared to be after the strap was broken. The instruction, if given, would have permitted conjecture by the jury that the fright might have occurred before the strap broke, and was the result of some noise or act not appearing in evidence, either directly or circumstantially. A responsible cause the weak strap-appears, which might reasonably have caused the fright, and, in due course, the injury. This is enough to sustain the verdict. To have granted the request would, it seems, have been error against plaintiff. But it is insisted that the verdict, standing at $2,000, is still excessive, as compensation for a temporary injury, though admittedly a violent and serious one. Here we have the judgment and discretion both of the jury and the trial court in favor of the lessened verdict. Appellate courts are always reluctant to act in such cases, and in no class of cases does the amount of damages rest so largely in the discretion of the trial court and jury as those for personal injuries. The judgment of the Circuit Court is affirmed. COMPTOGRAPH CO. v. MECHANICAL ACCOUNTANT CO. 1. PATENTS INFRINGEMENT. No. 612. Infringement is not avoided by the fact that the defendant's device performs some other function additional to that of the patented device. [Ed. Note. For cases in point, see vol. 38, Cent. Dig. Patents, § 380.] 2. SAME COMPUTING MACHINE. The Felt patent, No. 465,255, for a computing machine, claims 7 and 8, which cover a subtraction cut-off, are for a combination and not merely an aggregation of parts and disclose invention. Also held infringed. Appeal from the Circuit Court of the United States for the District of Rhode Island. lee. For opinion below, see 140 Fed. 136. John W. Munday and Henry Love Clarke, for appellant. Wilmarth H. Thurston (Warren R. Perce, on the brief), for appel Before COLT, PUTNAM, and LOWELL, Circuit Judges. COLT, Circuit Judge. This suit was brought for infringement of letters patent No. 465,255, dated December 15, 1891, granted to Dorr E. Felt, for improvements in computing machines. Computing machines are for performing arithmetical problems mechanically, thereby doing away with the mental operations incident to their performance in the ordinary way. Computing machines of the type of the patent in suit are simply mechanical adding machines. The alleged infringement is limited to claims 7 and 8 of the patent. These claims cover a single feature of the Felt machine, known as the "subtraction cut-off." This cut-off is a simple mechanical device for the correction of the error of 1 on the left of the true remainder, which always appears in additive subtraction when the complement of the subtrahend is added to the minuend. The device consist of a series of pivoted levers so arranged with respect to the numeral-wheels and carrying mechanisms as to throw out of operation, in any given sum in additive subtraction, the particular carrying-pawl which causes this error of 1. In adding machines prior to the Felt patent, when this error in the remainder appeared upon the reading line of the machine, it either had to be corrected on the machine or mentally disregarded. Felt was the first inventor to incorporate into an adding machine the mechanical means for the prevention of this error. The error in question may be illustrated by the following example: For instance, to subtract 6 from 9, we first find the complement of the subtrahend, 6, by subtracting 6 from 10, which gives us 4; we then add 4 to the minuend, 9, which gives us 13, or the correct result, 3, plus an error of 1 on the left. In performing this example upon adding machines of the Felt type, we would first strike the 9 key of the units series of keys, which would bring the figure 9 on the units numeral-wheel to the sight or reading line of the machine; we would then strike the 4 key of the same series, which would bring 13 to the reading line, the 3 being on the units numeral-wheel and the 1 on the tens numeral-wheel. By the operation of the Felt subtraction cut-off, this error of 1 on the tens numeral-wheel will be prevented, and only the true remainder, 3, will appear on the reading line of the machine. The cause of this error is manifest when we look into the nature of the operation of additive subtraction. We have, as illustrated in the foregoing example, performed subtraction by adding the complement of the subtrahend, 6, which is 10-6, or 4, to the minuend, 9, making the result 13, or 10 in excess of the true remainder. In this operation we have committed the obvious error of increasing the minuend by 4 instead of decreasing it by 6, and the measure of this error is plainly the sum of 6 and 4, or 10. In other words, the measure of the error is the power of ten from which the subtrahend is subtracted to produce the complement. It follows that the error in the remainder in any given example of additive subtraction is 10, or 100, or 1000, or 10,000, according to the power of 10 from which the subtrahend may be subtracted, and, since naughts count for nothing, that this error will be represented by 1 on the left of the true remainder. In referring to this feature of the Felt machine, the patent says: "Another object of my invention is to prevent the carrying of tens from any column to the next higher whenever a subtraction is made by means of adding a complementary number, as hereinafter explained." We have here the underlying conception of the Felt subtraction cut-off, which was so to organize his machine as to prevent the operation of the carrying mechanism whenever it is necessary to prevent it in performing additive subtraction. That the error in question is caused by the carrying mechanism, and may be prevented by preventing the operation of a particular carry, clearly appears from the following considerations: As the tens are carried mentally in arithmetical addition, so an adding machine in its normal operation carries the tens mechanically by carrying 1 to the numeral wheel of the next higher order, as from the units wheel to the tens wheel, and from the tens wheel to the hundreds wheel, and so on throughout the series of wheels. Now, it is apparent that this error of 1 on the left of the true remainder is due in each example to the particular carry of 1 either from the units column to the tens column, or from the tens column to the hundreds column, or from the hundreds column to the thousands column, dependent in each case upon the power of 10 from which the complement is subtracted. For instance, in the example already given of subtracting 6 from 9 by additive subtraction, the error of 1 in the tens column was caused by the carrying of 1 from the units column; and, if this carry had not taken place, the result would have been the true remainder, 3, on the reading line of the machine. So, if the example had been to subtract 66 from 99, the error of 1 in the hundreds column would have been⚫ caused by the carrying of 1 from the tens column to the hundreds column. So, if the example had been to subtract 666 from 999, the error of 1 in the thousands column would have been caused by the carry of 1 from the hundreds column to the thousands column. The patent then proceeds to point out the way this error had to be corrected in prior adding machines, including the machine covered by an earlier Felt patent. This correction was made on prior machines by striking once all the 9 keys to the left of the true remainder, and striking the last 9 key ten additional times. This operation was known as "running off the nines." For instance, in the example already given, the subtraction of 6 from 9, in which the remainder on the reading line is 13, the error of 1 in the tens column was corrected by first striking the 9 key of that column, which brought a 0 to the reading line of that wheel, but which at the same time, through the carrying mechanism, caused a 1 to appear on the hundreds wheel. This 1 must be got rid of in the same manner by striking the 9 key of that column, and so on through the entire series of wheels to the left. This would still leave the 1 on the extreme left-hand wheel, which has no series of keys attached to it. To cancel this 1 requires ten depressions of the 9 key of the last or adjoining series. The patent then proceeds to describe the means employed for preventing this error, the mode of operating these means, and the result accomplished: To save the carrying of the 1 above mentioned and afterwards adding the 9's as above described, the following device is provided: There is a series of levers, one for each carrying-pawl. Each lever is provided with a finger-piece, which extends upward within convenient reach of the operator. Each lever is also provided with an arm adapted to engage with the under side of its carrying-pawl and hold the carrying-pawl from engagement with its numeral-wheel. When the operator makes a subtraction, the lever at the left of the highest column or series of keys in which a key is struck in the subtrahend is pushed backward, which raises the carrying-pawl which carries from this highest series or column of keys to the next column to the left. By this construction and operation the carrying-pawl is held out of operation, so that the complementary number is added and the tens of the highest series are not carried, thereby making a proper subtraction. Having thus clearly set forth the object of his invention, the way in which the error in question was corrected in performing additive subtraction on prior adding machines, and the mechanical means he has devised for the prevention of this error, the patentee then proceeds to claim this invention in the following claims of the patent: "(7) The combination, with numeral-wheels, actuating devices therefor, carrying-pawls, and devices for actuating said carrying-pawls to carry the tens, of a device for preventing the operation of the carrying-pawls, whereby the operation of subtraction is accomplished, substantially as specified. “(8) The combination, with numeral-wheels, 4, actuating mechanism therefor, and carrying-pawls, 27, of levers, 44, having arms, 46, substantially as and for the purpose specified." These combination claims were intended to cover the mechanical device described in the patent for facilitating and perfecting the per formance of additive subtraction upon an adding machine. They were manifestly not designed to cover generally the operation of additive subtraction, since the patent expressly states that such problems had been performed on prior machines. Nor were they, when read in connection with the specification, designed to cover every device for preventing the operation of the carrying-pawls upon an adding machine. They plainly were intended to cover, and do in fact cover, as one of the elements of the combination, the subtraction cut-offs described in the patent. When claim 7 refers to this mechanism as “a device for preventing the operation of the carrying-pawls," these words must be read with what follows, namely, "whereby the operation of subtraction is accomplished, substantially as specified," that is, in the manner set forth in the patent. The terms of claim 8 are more specific, in that it mentions the "levers, 44, having arms, 46." The only proper and reasonable construction of both claims is that each is for the combination of elements for performing complementary subtraction upon an adding machine in the way described in the patent, that claim covers as one of its elements the series of pivoted levers known as the subtraction cut-offs, and that claim 8 covers as one of its elements such levers when provided with the "arms, 46." It is contended that these claims are for mere aggregations. This position, however, is obviously untenable, since the subtraction cutoffs, or pivoted levers, co-operate with the carrying-pawls in the production of a new and useful result. It is true that only one pivoted lever co-operates with one carrying-pawl in the performance of any given sum in additive subtraction. But it by no means follows from this circumstance that the claims are for mere aggregations, or even that the series of pivoted levers are mere aggregations. not dealing with a subtraction machine for performing a single sum, but with a machine for performing all like sums, and this problem could only be solved by the incorporation into such a machine of the means by which all such sums could be performed. The fact that only one pivoted lever happens to be called into operation in performing a given sum is immaterial. The whole series of levers are indispensable when we take into consideration the subject-matter with which we are dealing. It is further contended that the subtraction cut-offs do not exhibit any patentable subject of invention, in view of the prior art. A large part of the present record, as well as the briefs of counsel, is devoted to a discussion of this branch of the case, and especially to the discussion of the so-called universal cut-off as an anticipation of the Felt device. No prior patent contains any reference whatever to any kind of a subtraction cut-off. The so-called universal cut-off of the Kelso 1886 patent, the Shattuck & Thorn 1882 patent, and the Carroll 1876 patent, was for the sole purpose of throwing out of operation all the carrying mechanisms in order that the numeral wheels may be turned back to zero, and so cancel the reading line on the machine. If it had been found, as the defendant attempted to show, that the universal cut-off could be used upon an adding machine during the operation of addi |