Plaintiff sues the defendant to recover for a shortage on a carload of barley delivered defendant carrier at Bordulac, North Dakota, for transportation and delivery to a consignee in Superior, Wisconsin. Plaintiff makes proof of loss by evidence that 62,440 pounds, by weight, of barley was placed in the car for shipment, and, according to the state weighmaster’s certificate on delivery, but 57,480 pounds were received by the consignee; and plaintiff seeks to recover at the market price per bushel for the difference in weights, 4,960 pounds. Defendant offered no testimony; and at the close of the case the trial court concluded that the facts were parallel with those of Miller v. Northern P. R. Co. 18 N. D. 19, 118 N. W. 344, 19 Ann. Cas. 1215, and directed a verdict of dismissal. Plaintiff appeals.
The evidence shows that the barley was hauled by plaintiff’s employees to an elevator managed by one Johnson, at Bordulac. The barley was weighed at the elevator as it was received into the elevator, and again weighed out as it was loaded from the elevator into the car, the weights corresponding. At least twenty-one different weighing operations were necessary in loading the car, as but 3,000 pounds could be weighed at once in the hopper from which it was placed in the car. But Johnson’s testimony is positive as to weights. He testifies:
“I know the scale was all right. I tested the scales every once in .a while to find out whether it was working right, and I know this scale was in good working condition and right when I weighed this grain. I know how many pounds of barley I weighed and put into this car. The amount was 62,440 pounds; and I know that that was the correct amount of barley that I put into that car. Immediately after I loaded the car I went over and billed it out and got a bill of lading. I noticed the car had been sealed when I went over and got the bill of lading. I got this bill of lading immediately after loading the car.”
On cross-examination he testified he “did not represent Mr. Morris (plaintiff). The barley that I put into the car came out of the elevator. I ran it down into the hopper and from there on to the hopper *139scales, and from there it was weighed and conveyed into the car. Nobody assisted me to load the grain. I loaded it myself and did the whole of the transaction, the weighing and the loading. I had been working there at that time about two months. During the two months I was there I couldn’t say whether the scales had been inspected by ■either the state or county inspector. I don’t remember of any such inspection having been made. So that as far as the scales being correct is concerned it is simply my testimony. I was never more particular than the fact that these scales balanced. I am not an experienced inspector, but I am a good mechanic and know everything about them. I never had any experience in testing scales; was never in that line of business. The principal part of my examination of the scales was to see whether or not they were balanced. I weighed the grain all in one operation. I don’t mean the entire load ivas put on the hopper scale at one time; I couldn’t put more than 3,000 pounds ■on it at one time. There were about twenty different weighing operations. The figures that I had down at the time of those weighing ■operations, — I had them, and if I remember right I gave them to plaintiff. I haven’t the figures at this time. I say that the total weight was 62,440 pounds. This was a little more than a year ago that I weighed this grain. I gave the weights to plaintiff the same day. Since the day of weighing I have not had possession of the figures. As soon as the car was billed out and turned over to the railroad I turned the figures over to Morris and I was through with the transaction. That was the last I had to do with it. The operation of weighing and loading a car of grain was an ordinary transaction in my line of business. It was something that I was doing almost every day, and there was nothing different in this operation at that time than the usual operation of loading and weighing out a load of grain. I remained at that elevator for three months. During the time I was there I weighed out and shipped about fifty cars of grain.” On redirect examination he stated: “I knew that 62,440 pounds of barley was the actual number of pounds that I put into that car. I was the person who billed out this car of barley and the person who took the shipping bill from the railroad agent. I know of my own knowledge how much the figures showed at the time, both before and after I gave them to Morris.”
*140Plaintiff testifies that the barley “was weighed twice; weighed into the house, then weighed out into the car. I did not weigh it myself, but Johnson did. I was there when part of it was weighed. I know of my own knowledge what the grain weighed into the house and out of it. I know it by the slips that Johnson handed me. The slips both corresponded, when weighed in and weighed out.” Witness never saw the car after it left Bordulac. “The only reason I believe that there was a shortage or loss of grain, and which is the basis of my cause of action in this case, was because and is because the figures which I testified to and which the witness Johnson testified to did not correspond with the figures of the weighmaster in Wisconsin. The amount of shortage I claim in this case depends entirely upon the weighing of the grain in Bordulac and Wisconsin.” The original bill of lading is in evidence, identifying the car in number by the testimony of the Wisconsin weighmastePs certificate. The testimony of the assistant weighmaster, who issued the certificate, is in evidence. He testifies: “I am the assistant weighmaster who* weighed the car referred to in Exhibit A. To my own knowledge I know that the scale was in good condition and working order, and weighed correctly. I know this because it had been tested a short time before and found O. K. I know that I weighed correctly the ear described in Exhibit A, and know the number of pounds of barley shown by Exhibit A is the actual number of pounds of barley contained in said car at said time.” Exhibit A is the official weighmastePs certificate and is in evidence. Such is the testimony upon which a verdict was directed for the common carrier.
So far as the facts in this case are concerned the common carrier insured the delivery at destination of all the barley that it received from plaintiff for transportation. As is stated in Miller v. Northern P. R. Co. supra, 18 N. D. on page 19, 118 N. W. 344, 19 Ann. Cas. 1215, “a prima facie case is established by proof that the carrier received the goods for transportation and failed to deliver them safely. Conversely stated the rule is that, in order to make out a prima facie case, plaintiff must prove that the goods received by the common carrier were not all safely delivered.” Plaintiff must establish by a fair preponderance of the evidence that a portion of the grain received at Bordulac was not delivered at Superior. To make his prima facie case *141he has offered the evidence above narrated. Has he made a prima facie ease of loss in transit to the amount of the shortage or at all? It is clear that both parties have tried this action with full knowledge of the holding in Miller v. Northern P. R. Co., plaintiff apparently purposely omitting to make any proof of the condition of the car on its arrival at Superior, and defendant assuming that under the Miller Case, construed with Duncan v. Great Northern R. Co. 17 N. D. 610, 19 L.R.A.(N.S.) 952, 118 N. W. 826, that to make a prima facie case plaintiff must establish some facts corroborative of proof of loss besides and beyond mere proof of a difference between initial and terminal weights. In some respects the facts differentiate this case from the Miller Case, but some principles of law there discussed must be here reviewed and affirmed or disapproved.
The trial court has followed certain statements in Miller v. Northern P. E. Go., as appears from the statement of the court to the jury at the time of directing a verdict in favor of the defendant company. In that case the court weighed the evidence, and in discussing the proof of the alleged loss in transit held plaintiff had not made a prima facie case. Such question a trial court, and necessarily on appeal an appellate court, must determine as a matter of law, and thereupon either find that a verdict for recovery would not be supported by the evidence, or, on the contrary, must find the proof sufficient to sustain such a verdict, should one be rendered; in which event the question of loss under a substantial conflict of evidence is for the jury to determine. To determine such prima facie ease the court, in Miller v. Northern P. R. Co., considered and discussed the testimony, and in so doing the court was strictly within its province. But from the language of the opinion, especially that used in the opinion on the rehearing had, the holding may be easily misunderstood as announcing a rule to the effect that a prima facie proof of loss cannot be made by proof of the difference between the shipping and terminal weights. Such is not the law. And what was there said upon and concerning presumptions has evidently mislead the trial court in this case, as well as counsel for defendant company. It is our duty to avail of this first opportunity to explain the matter and announce a definite rule.
In speaking of the weights at place of shipment and at place of delivery, in Miller v. Northern P. R. Co. the following language was *142used: “Not a scintilla of evidence was offered aside from the above to» prove that any flax was in fact lost or removed from the car; plaintiff’s entire case resting upon the mere inference aforesaid. Thus it is seen that plaintiff’s entire case rests upon a mere inference or presumption based not upon facts, but merely upon other presumptions, to wit, that the Barlow scales were accurate, and that the public record in the state weighmaster’s office [at Duluth] speaks the truth. Presumptions or inferences cannot be based upon other presumptions, but must be based upon proven facts. Not only this, but a presumption ordinarily has no probative force, and when contrary evidence is adduced the presumption disappears.” Similar statements and reasoning appear at various places in that opinion. This is misleading. Proof of a certain weight is proof of a fact. Whether presumably correct does not matter. Evidence of weight is nevertheless evidence of the fact of an amount to be ascertained by such means. The proof that at shipment this barley weighed 62,440 pounds, prima facie establishes its quantity, the determination of which is essential to a recovery. Proof that at delivery but 57,460 pounds of barley was contained in the car establishes prima facie the amount delivered the consignee. Thus, proof is made of the fact of the amount of grain received and the amount delivered. The amount of loss may be termed a calculation, an inference, or a presumption; it does not matter what. But as it concerns a fact, to wit, a numerical difference in amounts, it must be a fact, inference of fact, or presumption-of fact. And as such are but matters of evidence they are for the consideration of the jury, and not ordinarily to be weighed by the court. As such they constitute evidence of the proof of loss in transit. While an inference might be drawn from the number or manner of taking of the weights, that either or both weights are inaccurate, such must be at most but an inference to be drawn by the jury, and not by the court, in weighing the testimony in the determination of the ultimate fact of whether more grain was received by the carrier than was delivered by it to the consignee. These must be but inferences concerning facts in evidence consisting of weights sworn to or established by the testimony; and the accuracy of the weights taken is for the jury to determine. ITence in discussing these questions it was error to assume th at there were any presumptions or inferences based upon other presumptions; so, also, the rule that *143presumptions cannot be based upon other presumptions can have no application. Presumptions at times are to be dealt with by the court; at other times presumptions are but part of the proof, and for the jury. In determining whether plaintiff has made a prima facie case, the courts must accept the facts or inferences of fact, and but apply the law. And such evidentiary facts to be taken by the court as established for such purposes consist of the evidence of the weights at Bordulae and Superior, which prima facie disclose a difference of 5,000 pounds. This difference is an established fact, not an inference or presumption, so far as the court is concerned. Had this grain been in sacks, instead of in bulk, and 500 sacks been delivered the carrier and 450 sacks been, by the carrier, delivered to the consignee, the inference of shortage ■would be the same as in this case, though the units or items from which the shortage is deduced would be different. Proof of shortage of 50 sacks might involve inferences of inaccuracy in the counting or other method of computation, but that would be a question of fact for the jury as in any case of conflict of testimony. This question of weights, it is true, involves added elements of uncertainty concerning the accuracy of the scales at both points, and their manner of use, with which the personal elements of honesty and credibility become more important possibly as the opportunity to defraud is thus increased; but all these questions must nevertheless remain questions of fact, and within the province of the jury, and without the province of the court, to determine. • Por an interesting discussion of presumptions and inferences, see chap. 13 of vol. 2 of Chamberlain’s Evidence; chap. 2, and especially § 104, of Jones on Evidence, 2d ed., from which section we quote: “Presumptions must be based upon facts, and not upon inferences or upon other presumptions. ‘No presumption can, with safety, be drawn from a presumption.’ The fact presumed should have a direct relation with the fact from which the presumption is dravra; but when the facts are established from which presumptions may be legitimately drawn, it is the province of the jury to deduce the presumption or inference of fact. If the connection is too remote or uncertain, it is the duty of the court to either exclude the testimony from which the presumption is sought to be deduced, or to instruct the jury that the evidence affords no proper foundation for any presumption. If, however, the facts are clearly *144established forming a proper basis for a presumption of law, the jury has no right to disregard the presumption which the law raises. The presumption in such case is one deriving its force from the law, and not merely from process of reasoning.” With initial and terminal weights shown as facts by competent evidence, the law draws therefrom the presumption of shortage for the purpose of carrying the case to the jury for their finding of the truth of the matter. Consult, also, §§ 15 to 48, inclusive, of Greenleaf on Evidence, 16th ed., in which latter section presumptions of fact are treated under the following summary: “This class of presumptions embraces all the connections and relations between the facts proved and the hypothesis stated and defended; whether they are mechanical and physical, or of a purely moral nature. It is that which prevails in the ordinary affairs of life, namely, the process of ascertaining one fact from the existence of another without the aid of any rule of law; and therefore it falls within the exclusive province of the jury who are bound to find according to the truth.” See also Wigmore on Evidence, vol. 4, §§ 2490-2494, from portions of which we quote: “And a ‘presumption of fact’in the usual sense is merely an improper term for the rational potency or probative value of the evidentiary fact regarded as not having this necessary legal consequence. ‘They are in truth but mere arguments,’ and ‘depend upon their own natural force and efficacy in generating belief or conviction in the mind.’ . . . There may be a preliminary question whether the evidence is relevant and admissible as having any probative value at all; but, once it is admitted, the probative strength of the evidence is for the jury to consider. . . . There is, in truth, but one kind of presumption, and the term ‘presumption of fact’ should be discarded as useless and confusing. Nevertheless it must be kept in mind that the peculiar effect of a presumption ‘of law’ (that is, the real presumption) is merely to invoke a rule of law compelling the jury to reach a conclusion in the absence of evidence to the contrary from the opponent. If the opponent does offer evidence to the contrary (sufficient to satisfy the judge’s requirement of some evidence), the presumption disappears as a rule of law, and the case is in the jury’s hands, free from any rule.” Any so-called presumptions or inferences deducible from the difference in weights must be of fact, and not of law; and to be finally drawn *145by tbe jury, and' not by tbe court. Tbe proof made of tbe weights ■either establishes evidentiary facts, or it does not, according as the jury determine the truth of the matter. But if such facts be taken as true, as they must be taken by the court, plaintiff has established his prima facie ease upon which he was entitled to have his case submitted to the jury.
If a difference between the shipper’s and consignee’s weights, approximating 5,000 pounds, on a carload of 62,000 pounds, where the testimony is positive as to the manner of the taking of the weights, and with the double weighing in and out of the elevator at the initial point as here had, did not make prima facie proof of loss in transit, then it is difficult to see how a distinction in principle could be drawn had the difference in weights been 60,000 pounds, instead of 5,000. If a court is to pass upon this matter as a question of law, where between those limits would the line be drawn at which the case would be regarded as involving a question of fact for the jury’s determination? If it can be said that upon mere weights alone no proof of loss can ever be made, why permit evidence of weight to be received? But if it be said that weights have some evidentiary force, but not sufficient of themselves to ever constitute proof ■of loss, then the question arises: Of what must the corroborating evidence consist that the law will recognize as supplementing the weights sufficiently to make, when considered with the weights, a prima facie case of shortage? And thus we are left to conjecture. This but illustrates the sound rule, that the whole question is one of fact or inference of fact from evidence, and for the jury to determine from the evidence, whether that may consist of weight's alone, or of added circumstances corroborative of shortage. And if proof of differing weights entitles plaintiff to go to the jury, if likewise makes proof of a prima facie case upon which, where the testimony, as here, is not conflicting and reasonably certain, plaintiff would be entitled to :a finding of the jury upon the question of loss; and if so, the amount thereof; under an instruction from.the court that the jury might consider, with the evidence, matters of which judicial notice might be taken, including the possibility of mistake in weights at either end, the interest, if any, of the witnesses in the subject-matter of the suit, .natural shrinkage or loss of weight in handling, and any other facts *146known or that might be proven which might enter in to explain any discrepancy in weights between the shipping point and destination; from all of which, as a question of fact, the jury must determine whether any less grain was delivered the consignee than was received from the shipper by the common carrier for transportation, and the amount of any such loss.
The only parallel case from another state the writer has been able to find is that of Schott v. Swan, 21 S. D. 639, 114 N. W. 1005, apparently not as strong a case for the shipper as the one before us, because of inaccurate methods disclosed in the evidence of the shipper in arriving at shipping weights. In that case a verdict for the shipper, based upon the bulk weight, though taken by a method that must have been approximate and inexact, was permitted to stand over the testimony of the sworn weights of the grain as it was taken from the car. The court held the question to be one for the jury on the theory of a conflict of evidence.
In view of evidence that may be presented upon a retrial we will say that should it appear, in addition to the facts in evidence, that upon arrival in Superior the car was in good condition, by inference negativing the escape in transit of any grain from the car because of defects in the car, and that on arrival the car was sealed, by inference causing the conclusion that it had not been opened during transit, still the question of whether any of its contents was lost in transit must remain a question for the jury. It may be that to a certain extent the question of loss is thereby left to conjecture, but that must be true to a certain extent in most verdicts. To here hold otherwise would result in exonerating the carrier as a matter of law, even though on arrival this car should be found to have contained but 10,000 pounds of grain, notwithstanding proof was made that over 60,000 pounds was shipped from Bordulac. In that event to assume such a gross variance between the testimony of the shipper and the consignee ■.as to weights to be attributable to a difference in weights only would clearly be an assumption of fact and an invasion by the court of the province of the jury. So, too, must such an assumption, only in lesser degree apparent in this case, be the same in principle. The same rule of law must apply in either and in all cases. .And in either case the jury, and not the court, must determine the ultimate fact of loss in *147transit. And to announce the rule that the degree of proof offered in this case as to initial and terminal weights does not establish a prima facie case sufficient to invoke the jury’s verdict, or that it may be overcome as a matter of law by mere presumptions or inferences of fact to be deduced from the condition of the car on arrival, would be to require of the shipper such a degree of proof of loss as is ordinarily impossible for him to obtain and produce, and announce a rule which will in effect abrogate the common carrier’s legal obligation as usually an insurer of property to be transported. Accordingly, where the proof is reasonably certain and satisfactory that grain of a certain quantity, as previously ascertained by its weight, is loaded in bulk in a car for transportation, and then is taken in charge by the common carrier for delivery to a consignee, and proof of the amount of such grain subsequently received by the consignee is made by evidence consisting of a state weighmaster’s official certificate, with initial and terminal weights, as so established, differing several thousand pounds on a carload of grain transported, plaintiff is entitled to a finding by the jury as to' whether any of the grain has been lost in transit, and, if so, the amount of such loss.
In view of another trial we do not deem it necessary to pass upon a question raised upon the measure of damages.
The judgment appealed from is therefore set aside, and this case remanded for new trial. Appellant will recover costs on appeal.