On March 13, 1953, plaintiff-appellant was one of a two man team repairing freight cars in the Union Pacific Railroad Company shops at its east yard near Los Angeles. At that time he was a regular employee of the company. On the particular morning the duties of the two required that they do some welding which involved the use of heavy duty electrical equipment. Upon completion of one welding job the plaintiff went alone to a large pole to disengage a rubber cable leading from a socket on the pole, the cable containing the electric wires which ran from the socket to the welding machine. The main length of the cable is the size of an ordinary garden hose. On the end of the cable making the connection at the socket, the cable expands into sort of a bulbous object about four inches in diameter. This bulbous end or plug fits into the socket on the pole. The underlying mechanical principle of the heavy industrial socket and plug used here is not unlike the ordinary household wall socket and appliance cord with plug.

While disengaging the cable with its plug from the socket on the pole, Conner claims that he sustained a back injury and that the injury was due to the socket being maintained on the pole in a negligent condition by Union Pacific. Conner sued the railroad company under the Federal Employers’ Liability Act, 45 U.S.C.A. § 51 et seq. The alleged negligence will be alluded to later.

The cable and plug, the socket, and a switchbox assembly immediately above the socket are in evidence, all attached to a section of a pole or post. The parties have only partially described in words the apparatus here involved.

Although repetitious, it may be said the major component parts (from top to bottom) would be as follows:

1. The switchbox with switch apparatus inside the box.

2. The casting containing the socket immediately below and connected to the switchbox.

3. The bulbous hard rubber covered plug which is securely attached to the garden hose-like cable.

The switchbox is rectangular in shape and inside is the heavy duty switch fastened to the back of the box. The switch-box has a door which takes up almost the whole face of the box. The door when opened swings upward. Inside the door of the box is a removable panel front which when screwed in place leaves only the switch handle within view. Otherwise, this front panel covers everything inside the box.

At the metal casting below the switch-box, called the socket, we find the critical portion of this case. This socket is bent pear-shaped, the neck of the pear being connected to the bottom of the switch-box. The lower or wider end of the bent pear bends down and away from the pole at about a 45 degree angle. Into this lower end of the metal socket the connection is made with plug and attached cable.

The pole being round, a board is affixed to the pole and the switchbox is fastened to the board. The narrow part or the neck of the pear-shaped socket is flat on the back and has two wing-like projections. By a screw through each wing of the socket, the socket is affixed to a board immediately below the board on which the switchbox rests. (On the day in question, there is no suggestion that the backboards did not remain firm to the pole. Likewise, it may be assumed that the switchbox remained firm against its backboard.) A connection between the switchbox and the pear-shaped socket casting is made by means of a threaded pipe or nipple. Around this nipple between the socket casting and the bottom shelf of the switchbox are two locknuts, one above the other. Around the nipple which protrudes less than an inch into the switchbox is a locknut, and above the locknut, another nut called a collar. Properly set up, the pear-shaped casting is tight against its backboard. The locknuts and collar pull the casting firmly up against the bottom of the switch-box.

As to plaintiff's injuries, the parties are not in agreement as to how bad his back now is, or has been. Union Pacific does not admit the plaintiff’s disconnecting the socket had anything to do with the serious condition of plaintiff’s back, or the pain he has suffered.

It may be that the plug end on the cable leading to the welder can be pulled straight out from the stationary metal portion of the socket, but it seems quite natural and much easier to wiggle the plug out, not unlike pulling a household plug out of a wall socket.

By his complaint plaintiff specifically charges only that the wings of the metal socket pulled over the screws of the backboard. And the reason for this was negligence in the maintenance of the metal socket with its attendant screws. No other negligence is charged.

Plaintiff testified that while disengaging the plug from the metal encased socket “just everything gave away” and instantaneously he had a sharp pain in his back. Other than that the plug disengaged from the socket, he only particularized that the wings of the socket came over the heads of the screws holding it to the backboard. There was corroboration of his contention that the company had fastened or maintained the socket to the board with too small or defective, rusty screws. No one directly or positively testified, and it was not pleaded, that the connection between the socket and the switchbox was loose on the day in question.

Possibly, if the jury had found in Conner’s favor there would have been enough evidence to sustain the verdict; that is, that the socket was negligently fastened to or maintained on its backboard and that the socket casting coming loose was the proximate cause of wrenching of Conner’s spine which activated serious latent trouble. For discussion here, it is assumed that Conner made a prima facie case. But the jury found against Conner which is not the customary thing in this kind of a ease.

Conner has appealed. He has two points. First, he complains of the district court’s stock instruction given to the usual effect that accidents can just happen without being the fault of anyone. Rarely is there a negligence case which does not justify this instruction. A careful perusal of the record indicates the giving of the instruction was justified here.

The second point is not so easy. During the course of the trial, one or two jurors did some questioning of the witnesses and so did the trial judge, these obviously being questioners against whom a party to an action finds it difficult to defend.

A juror thought there were signs of fresh or recent wrench marks, presumably made by someone putting a wrench inside the switchbox and tightening the locknut and collar against the bottom of the switchbox where the nipple comes through and is grasped by the locknut and collar. If the connection between the metal socket and the switchbox were loose when Conner detached the plug and the socket wings pulled over the screws, naturally there would have been more play in the socket and thus maybe more negligence.

As plaintiff’s attorney seized upon the “manna” that came from the juror’s and the judge’s questioning, the judge ruled plaintiff could not have it. The judge held that plaintiff himself had introduced into evidence the section of pole with the switchbox, socket, plug and cable (brought to the courtroom by defendant) and that plaintiff could not impeach the assembly. (The connection in the bottom of the switchbox was firm the day of trial.) And during argument to the jury, when plaintiff’s attorney was telling the jury that Union Pacific quite obviously had recently tightened the connection, the court remonstrated that plaintiff had introduced the apparatus and was bound by it.

The exhibit, consisting of the pole section, with switchbox, socket and plug seems to have just “backed into” evidence without any real foundation. From the record it appears that the exhibit came into evidence in the beginning apparently for the purpose of illustration.

Defendant did not pretend at the trial that the wings of the socket now have the same screws in them as on the day of the accident. Now the socket is screwed firmly against the backboard. The whole apparatus coming in for illustration, it does not appear (the plaintiff never having represented identity of condition) that the court was correct under the circumstances in its view that as a matter of law plaintiff had to vouch for the whole assembly. To illustrate: as indicated, the trial judge freely let the parties show the condition of the screws on the wings of the socket had been changed two or three times since the day of plaintiff’s injury. The same thing ought to apply on testimony about the connection of the socket with the switchbox. But an objection could have been made that a loose collar on the socket was not within the pleading. While temporarily a good objection, doubtless this point might have been met by an amendment to the complaint.

But assuming a loose collar at the connection of switchbox to socket on the day of the accident was material to the issues, if the court’s ruling was wrong on the ground stated, it was harmless error.

The cuts or deep scratches in the metal of the collar and locknut seem to be subject to an inference that they are wrench marks. And in color they do appear lighter and fresher than the rather weathered look which the collar and locknut inside the box as a whole bear. But assuming the connection within the box has wrench marks, there is no showing of the history of the locknut and collar being affected by these wrench marks. Who could say whether the scratches on the locknut and collar inside the switchbox, protected from the weather in the period from the injury to the time of the trial, were received from a twisting with a Union Pacific wrench last week, last month, or last year. But assume it can be safely inferred that Union Pacific after the injury and before the trial did tighten the collar and locknut inside the switchbox. Then where is the evidence of how loose the connection was on the day plaintiff says he was injured? Were they tightened just a fraction of a turn or several full turns ?

So, if the trial court was wrong in saying the plaintiff was bound by the present condition of the connection between the socket and switchbox, error, if it be, was harmless because the plaintiff proffered no competent evidence of loose collar or locknut on the day in question.

Generally, a condition shown to exist on a given day may sustain an inference as to identity of condition at a time backward or forward. Wigmore on Evidence, 3d Ed., Sec. 437. But whether such evidence is reliable enough for use is usually for the sound discretion of the trial court. With the paucity of foundation here, it would not have been permissible for Conner’s jury to draw any inferences from the wrench marks in the bottom of the switchbox at points on the collar or locknut. On the state of the record at the time the court said the plaintiff was bound by the exhibit the plaintiff was just not entitled to capitalize the wrench marks. See Ivinson v. Hutton, 119 U.S. 604 , 7 S.Ct. 403 , 30 L.Ed. 509 ; Hanover Fire Insurance Co. v. Merchants’ Transportation Co., 9 Cir., 15 F.2d 946 ; Shanahan v. Southern Pacific Co., 9 Cir., 188 F.2d 564 .

The judgment for defendant is affirmed.

. Perhaps it would be correct to refer to the whole — the casting with its recessed mouth to receive the plug and to the plug — as “the socket.” In an attempt to achieve clarity, herein the casting is referred to as the socket, and the bulbous end of the cable, which with the cable detaches from the casting, is referred to as “the plug.”