Maltbie, J.

The trial court has found that the injury to the decedent was caused by a blow received by his coming in contact with the rod attached to the end of one of the coal chutes, and that this rod had fallen from its place and swung over the track because it was not properly held in place by reason of the defective condition of the catch which ordinarily supported it.

The first claim pressed upon the attention of the court is that such a finding is not supported by the allegations of the complaint, and therefore cannot serve as the basis of a judgment. The fifth paragraph of the complaint alleges, among other things, that-the death of the plaintiff’s intestate was caused by the negligence of the defendant "in failing to provide the plaintiff’s intestate with a safe place in which to work.” The rod, when not in use, would, unless provision was otherwise made for it, hang over the track in a position dangerous to the employees of the company in the pursuit of their ordinary duties. The provision of a catch to hold it out of the way was clearly an incident to the duty of the defendant to furnish its employees with a safe place in which to work; and, having provided an adequate catch, the obligation to maintain it in proper condition rested *307 upon the defendant as part and parcel of the same general duty. Rincicotti v. O’Brien Contracting Co., 77 Conn. 617, 620, 60 Atl. 115. Negligence in the furnishing of an inadequate catch in the first instance, or in the failure to maintain in proper condition, by inspection and repair, a catch which when furnished was adequate, alike fall within the allegation quoted from the complaint. . Having gone to trial upon the merits of the case without raising any question as to the generality of this allegation, the defendant cannot now object to it. Eckert v. Levinson, 91 Conn. 338, 340, 99 Atl. 699; Cargan v. Harris, 90 Conn. 188, 191, 96 Atl. 940.

We are not greatly impressed with the contention of the defendant, that it was in fact misled by the pleadings and the evidence offered by the plaintiff, and did not anticipate that the question of its liability might hinge upon its performance of its duty with reference to the support of this rod. In the cross-examination of one of the plaintiff’s witnesses, it called out a description of the rod and of the method of its support when not in use, and the fact that, if not properly secured upon its catch, it would hang over the track; and in arguing its motion for a nonsuit at the close of plaintiff’s evidence, it made distinct reference to a possible claim of the plaintiff that it had failed to provide the deceased with a safe place in which to work. If the defendant saw fit to leave unchallenged the broad allegation of the complaint, it was its duty to prepare its case to meet any possible claim that might be made under it.

The trial court has distinctly based its conclusion upon the application of the doctrine res ipsa loquitur to the facts found by it to have been proven by the testimony. The defendant contends that these facts are an insufficient basis for the doctrine, in that there is no finding of the “time, place and manner” of the oc *308 currence of the injury. Upon the evidence it would doubtless be difficult to determine the precise time of the accident, and the particular chute which was its cause, and perhaps the court might hesitate to determine the engine upon which the deceased was then riding. But it is difficult to see how a lack of particularity in any of these respects could invalidate the court’s conclusion. It is true that one of the elements necessary to that conclusion was the fact that the particular chute in question had not been used by the deceased on the day in question; but that fact the court does find; and it is the necessary result of its finding that the injury did not occur at chute No. 7, where the engine upon which the deceased was then employed had been coaled, and, also, the fact, supported by fair inference and not questioned by the defendant, that no other chute than No. 7 had been used by the deceased on the night of his injury. The trial court has found these essential facts: the engine upon which the deceased was working, having been coaled at one of the pockets, was being backed past several of the others; while being so backed, the deceased was pushed against and struck by the rod which was attached to one of those other chut.es, and which, instead of being secured in its proper place, was hanging over the track.

If one examines the testimony and the exhibits in the light of common sense and his own practical experience, one finds in them reasonable support for every element there included. It is true that the testimony given as to the statements of the deceased in telling how his injury occurred, would, on its face, indicate that the injury was due to one of the hangers rather than to one of the rods; but these statements were, in all but one instance, made to persons who were evidently unfamiliar with the operation of the chutes and who might well fail to distinguish between the rods *309 hanging from the chutes, and the irons called distinctively “hangers ”; and, in the other instance, were made in English, a language which the deceased spoke very poorly. Against the prima facie effect of these statements the trial court might well have placed the probabilities arising out of the nature of the injury suffered, the relative positions of the rods and the hangers with reference to the chutes and the manner in which they were respectively fastened to it, as well as the evidence tending to prove that when the deceased came on duty that night there were no hangers on the chutes,, that other engines had been run back and forth past the chutes before the one on which the deceased was injured, and that only one chute had been used that night. To hold unwarranted the trial court’s conclusion that the injury was due to a blow from one of the rods rather than from a hanger, would be to invade its right to determine the facts in the case.

But the defendant contends that the trial court in the main arrived at its conclusions of fact, not from direct testimony, but by inference, and that to apply to conclusions so arrived at the res ipsa loquitur doctrine is to build inference upon inference in a way which the law will not permit. It is true that the trial court’s conclusions as to the manner of the accident were largely by way of inference; and it is also true that the doctrine res ipsa loquitur is nothing more than a label for the principle that, in certain cases, a court may, from the proven facts, infer the ultimate fact of negligence. Stebel v. Connecticut Co., 90 Conn. 24, 26, 96 Atl. 171; Thorson v. Groton & Stonington Street Ry. Co., 85 Conn. 11, 15, 81 Atl. 1024; Zeigler v. Danbury & Norwalk R. Co., 52 Conn. 543, 554; Button v. Frink, 51 Conn. 342, 347. But we cannot grant that the trial court committed error in taking these successive steps to its conclusion. In State v. Kelly, 77 Conn. 266, 271, 58 Atl. 705, *310 this court said: “Evidence for the purpose of creating an inference of a fact upon which to base an inference of another fact is generally inadmissible. It is too remote.” And in Levidow v. Starin, 77 Conn. 600, 603, 60 Atl. 123, this court said that a court is not required to promote the construction of “a cob-house of inferences.” The guarded language of these decisions points the way to the true rule. To hold that an inference may never be rested upon an inference, would be to do violence to many decisions of this court. Thus, motive is relevant in a criminal case only as affording a basis of an inference of guilt from its presence, or of innocence from its absence; yet “motive is a fact which may be inferred from circumstances.” State v. Buonomo, 88 Conn. 177, 184, 90 Atl. 225; State v. Saxon, 87 Conn. 5, 86 Atl. 590. So, on the issue of the mental capacity of a testator, words and acts prior and subsequent to the execution of the will are admissible, that from them his mental capacity at the time of their occurrence may be inferred, and from his condition then, a further inference may be drawn as to his condition when the will was executed. Cullum v. Colwell, 85 Conn. 459, 465, 83 Atl. 695. And for other similar illustrations of inferences resting upon inferences, see Spencer’s Appeal, 77 Conn. 638, 60 Atl. 289; Moffitt v. Connecticut Co., 86 Conn. 527, 86 Atl. 16; State v. Williams, 90 Conn. 126, 130, 96 Atl. 370. There is, in fact, no rule of law that forbids the resting of one inference upon facts whose determination is the result of other inferences. 1 Wigmore, Evidence, §41; State v. Fiore, 85 N. J. L. 311, 320, 88 Atl. 1039, 1042. It is but a rule of caution; its true function is to guide the court in the exercise of its judgment in determining whether or not evidence offered is too remote; Wynehouse v. Mandelson, 84 Conn. 613, 617, 80 Atl. 706; or, in making its final decision, in deciding whether the plaintiff has established a reason *311 able probability. Hoyt v. Danbury, 69 Conn. 341, 348, 37 Atl. 1051.

The reasons of appeal which have not been, in effect at least, already answered, do not merit discussion. No ground exists for correcting the finding; and the trial court was justified in holding that the conditions specified in Stebel v. Connecticut Co., 90 Conn. 24, 96 Atl. 171, as requisite for. the application of the doctrine res ipsa loquitur, had been fulfilled.

There is no error.

In this opinion the other judges concurred.