Plaintiff brought this action to recover damages for the death of her husband while in the employ of defendant, occasioned by inhaling the vapor of ammonia, alleged to have occurred through defendant’s negligence. Defendant owned and operated a brewery at the city of Detroit, and, in connection therewith, on March 10, 1902, operated a refrigerating plant, using a system of pumps, coils, and pipes, through which liquid ammonia was forced for refrigerating purposes. The entire refrigerating plant, including the pumps, pipes, joints, and gaskets all complete, was made in Detroit, and was installed in defendant’s brewery in 1898. About 1,000 pounds of anhydrous ammonia was used to charge the refrigerating machine. From the machine the ammonia is pumped through pipes and coils through the cellar, is compressed, expanded, and then cooled and returned to the machine to be used again. These pipes when they leave the machine are three-quarters of an inch in diameter and run from the engine room into the *373cellar. As they pass into the cellar, is placed an expansion cock three-quarters of an inch in diameter, and beyond this expansion cock the pipes are two inches in diameter. This expansion cock is barely opened from its seat, so that the point of a needle can just about be put through it. At the time the plaintiff was injured, there was a pressure of 165 pounds to the square inch on the pipe between the machine and the expansion cock, and a pressure of about 25 pounds to the square inch on the farther side of the expansion cock. Henry Heuge was defendant’s engineer, and had charge of the refrigerator plant, including the machinery, pipes, and coils generally, and the gaskets.
Plaintiff’s intestate was employed at general work around the brewery, and about 11 o’clock a. m., on March 10, 1902, he was cleaning out some beer tanks in the cellar when a rubber gasket blew out from between the flanges of a joint located behind the door leading from the cellar into the engine room, and the ammonia vapor escaped from the pipes. The joint where the gasket blew out was a flange joint. The gasket fits into a circular depression or channel in the nature of a countersink in the flange; the other flange having a corresponding projection from the face of the flange fitting quite closely the depression in the former, thereof and thereby inclosing and including the gasket within iron walls when the four bolts at the corners are screwed up tight. '
The declaration charged that plaintiff’s intestate was injured by reason of defendant’s negligence in failing to provide a safe place to work, in failing to keep its plant in a safe condition; in failing to use lead gaskets, in using rubber gaskets, and in failing to properly inspect the pipes and gaskets, and in using the same without any proper inspection. At the close of plaintiff’s testimony, the court directed a verdict for defendant, on the ground that no negligence had been shown on the part of defendant. Plaintiff assigns error upon the direction of the verdict and upon rulings of the court as to the admissibility of *374testimony. Did plaintiff establish a prima facie case of negligence on the part of defendant ?
Henry Heuge testified that in about five minutes after Kaéss came out, he went into the cellar and found that the gasket between the flanges of a certain joint behind the door between the cellar and the engine room was blown out. It was a rubber gasket about two inches in diameter on the outside, and the joint was a flange joint —an ordinary iron flange.
“I did not examine the condition of this gasket that had been blown out. I took a knife and scraped it off and put in a new one and bolted it up.
“Q. How much did you scrape off?
“A. About half or a little over.
“ Q. The rest was blown out ?
“A. It must have been somewhere. I couldn’t find it.
“Q. Did you see it ?
“A. No.”
This joint was so located that the door referred to, if opened wide, would strike it; but there was no evidence that the door ever had struck the joint.
“When the gasket is put into the female and the male is on top of it and screwed up tight, that flattens the gasket that is put between them. The idea of having the gasket between them is to make a joint tight, and to keep up the pressure. If the pipe were put together with a gasket which was not tight and the machine started, the ammonia would leak through the crevice where it did not fit tight, and the object of the gasket between the male and the female is to stop up the opening and make everything tight. Both rubber and lead gaskets aro used. Practically more rubber than lead. * * *
“Q. Did you ever make any inspection of these gaskets or these joints ?
“A. No, sir. * * * I walked through there and looked around to see if everything was safe and tight. That is all the inspection I made.”
Rubber gaskets and lead gaskets were used interchangeably, some preferring one and some another. The witness had been in the cellar about 10 minutes before Mr. *375Kaess was injured. When he scraped off the gasket, it felt soft to the point of his knife, which might have been occasioned by a little oil getting in there.
“We have to keep the cylinder oiled when it is under such high pressure. It is pretty hot when the compresser presses the ammonia. * * *
“Q. What is the effect of the oil left in the ammonia and circulating through the pipes upon that rubber gasket ?
“A. It will soften it on the inside. It will soften it.”
Oscar Lamsens, a witness sworn on behalf of .the plaintiff, testified: That he had been connected with the Stroh and the American breweries, and was connected with the Tivoli brewery from October, 1901, to April, 1902, during which time he took the place of the brewmaster while he was absent in Europe.
“ During my experience in different breweries, I have become acquainted and familiar with the process of using ammonia in the refrigerating machines. I have had practical experience for the last 3% years, and have had a technical knowledge for the past 10 years. I acquired my technical knowledge in a brewing school in connection with the university at Berlin, which I attended for two years. They used refrigerating machines there, and I studied them. I had practical experience with the refrigerating machines at the Stroh, the Tivoli, and the American breweries.”
That a special oil is always used. They pump it into the system, and it goes over the cylinder and into the oil trap. The rest of it-goes into the system and comes back.
“ Q. What is the effect of the oil upon rubber used for making rubber gaskets ?
“A. I think as long as the gasket is tight between the two flanges, it could not affect it at all. * * *
“ Q. Do you know what kind of gaskets were ordinarily used in joints in' systems connected with a refrigerating pla^t at f^e time of the accident ?
“A. Wi only used lead; but I know some other firms that use rubber. * * *
*376“ The Court: You have never seen the effect of rubber or lead gaskets except at the American P
“ A. JNo, sir.
“ The Court: Do you use any rubber out there ?
‘-‘A. No.
“Q. Then you don’t know the effect upon a rubber gasket when it is used upon the joints of an ice machine ?
“A. No. * * * Ammonia is dangerous to handle, and the danger increases with the increase of the pressure. The purpose of the bolts in the" holes of these flanges is to hold them and to tighten them up. It is necessary to have the joints tight. If a joint was made with the male and the female joints and the bolts were to be loose, the joint would leak. If the flange is tight, it does not leak. If the gasket is properly put in and the bolts are tight, it will not leak.
“ The Court: Do you know the character of ammonia, whether it is penetrating or not, ammonia fumes ? If this was not absolutely tight, would not the ammonia leak out ?
“A. Yes, sir.
“ The Court: In other words, if it was not absolutely tight, then there would be sufficient leakage so that it would indicate a leakage without danger to life.
“A. Yes, sir; I could smell it.
“ The Court: In other words, you would discover it by the smell ?
“A. Yes, sir; I could smell the leak.
“ Q. Would you consider that a gasket would blow out from a small leak ?
‘ ‘A. It would be a big leak. The rubber adheres to the face of the flange when it is screwed up. It sticks on there.
“ Q. How is it removed when they break the joint and take it out ?
“A. With a knife.
“ Q. Does it cling hard ?
“A. Generally you go around with a knife and clean it off with a knife.
“ Q. When a gasket is once put in the joint and it is tightened up, what, if anything, is necessary to do with the bolts afterwards to keep it tight ?
“A. When the gasket is once tight, it generally stays tight, unless by changes in the temperature or something of that kind it becomes loosened. The higher the temperature, the more it will loosen up. If it loosens up, you tighten it.
*377‘ ‘ Q. What do you call following up a bolt ?
“A. Taking them up, one after the other, tightening them up, one at a time. If you have a joint tight in the first place, you do not go over it until you smell a leak. It is not necessary.
“ Q. Could a lead gasket blow out of that joint such as there was there, if it was tight ?
“A. I know that the lead gasket would not blow out, in my judgment.
“ The Court: Have you ever seen an instance where a gasket has blown out ?
“A. That is the only one.
“ The Court: The only one you ever heard of ?
“A. Yes, sir.
“ Q. As I understand it, the inspection you made of these machines when they are tight, if the joint is tight, you go through there and nothing is done with the joint unless you smell ammonia ?
“A. Yes, sir. * * * Ammonia, when it comes under high pressure, is liquid and is compressed, and, when it is released, it vaporizes instantly, and forms fumes. If it were turned into this room, it would go all over, and that property of expansion is very fast; awfully fast. * * *
“ Q. You never examined the joint, did you?
“A. No, sir.
“ Q. Did you examine it after it was blown out ?
“A. No, sir.
“ Q. So far as you know, it had never leaked ?
“A. No, sir.
“Q. And the fact that it had not leaked would be an indication to cause you or any engineer to think that that joint was in proper condition ?
“A. Yes, sir'.
“ Q. The first signal of danger that you would get would be the smell of the liquid ?
“A. Yes, sir.
“ Q. And that experience would apply to all machines ?
“A. Yes, sir.
“ Q. And you consider that a signal ?
“A. Yes, sir. * * *
“ Q. I will ask this question: If that joint had been screwed tight, and the gasket had blown out with 165 pounds pressure on it, could it have blown out with 165 pounds on it, if the joint had been screwed up tight ?
“A. I don’t know the condition of the gasket. * * * *378The flange might have something to do with it. If the female did not fit on the male, it would not be possible to make a tight joint.
“The Court: If that joint was drawn up tight with a rubber gasket in it, what pressure would blow it out, if you know ?
“A. I cannot tell you that. * * *
“Q. Can you tell approximately ?
“A. Yes, I judge 200 pounds; not over that.
“Q. 200 pounds would not blow it out ?
“A. No, sir; it would not blow it out. * * *'
“Q. Now, I ask you when the plant is installed and properly set up, whether or not the inspection that is made by the person who is running the plant to ascertain whether or not these joints need attention, comes from a small leakage or small ammonia gas and from a leak in the joint — isn’t it true that that is the only time that you give them any attention ?
“A. Yes, sir. We screw them up then. * * *
There is no plant that don’t have some leaks. A hole was punched through one of the pipes down there. It was quite small. We did not determine how much of a hole it was, but if you could put the smallest needle into it, it would be enough to fill the room with ammonia inside of a minute. ”
Page Eaton, a witness for plaintiff, testified:
“ I looked at' the place where the ammonia came through the pipe as soon as the ammonia got out so that we could get in there. I should judge that was about 15 minutes after the accident. Before that time the engineer had a hose in the cellar to scatter the water and to condense the ammonia. In the place where the ammonia came through, the gasket was blown out. This was behind the door. A little past the door. * * * I saw a piece of the gasket that blew out. It was on the floor, I believe, near the joint. The engineer had it, if I remember right, when I saw it. He showed it to me, we were looking at it. It seemed to be rotten and old. I don’t just remember how it did look.”
Charles Tart, a witness for plaintiff, testified:
“Q. Now, will you state if that joint which blew out was made with a rubber gasket such as this joint is, with a rubber gasket inside of it and having bolts which would *379go into the four corners and which were screwed up tight, whether that gasket could have blown out ?
“A. No, sir.
“Q. At 165 pressure ?
“A. No, sir.”
In the course of the trial, plaintiff’s counsel stated his theory of the case, as follows:
“My theory is that the door had been striking this joint for four years, and had loosened it, and the oil run down through there and rotted it, and a rubber gasket is an improper gasket to use in that place. I intend to show that if this joint was properly made, it was impossible for it to have broken, regardless of whether rubber or lead was used. I will not leave it to any inferences.”
At the close of the testimony the following occurred:
“The Court: It is your claim that the accident happened because the flange was not screwed up tight ?
“Mr. Cook: Yes, it was not screwed up tight. It sets in there three-sixteenths of an inch, and if it had been screwed up tight, the gasket could not have come out. It would be up against the face of the flange, and it could not have blown out unless the flange had been separated somewhat. My claim is that this joint became loosened in some way, and that no inspection had been made to discover the condition, and that is negligence.
iiThe Court: Then you make no claim that by reason of rubber gaskets there was negligence ?
“Mr. Cook: I will admit that the testimony does not show that.”
No requests to charge, so far as the record discloses, were presented by either side. In the course of the charge, the court said to the jury:
“ The attorney for plaintiff withdraws any claim that there was negligence by reason of an imperfect gasket, and plants his case upon the theory that this accident could not have happened unless the bolts which held the flanges together were loose, and that the duty which the defendant owes the plaintiff under those circumstances was an inspection of those things which would reveal the loose condition.”
It is claimed by counsel for plaintiff that this was an *380incorrect statement of his position, and that the effect of his statement with reference to the gaskets was to admit that the testimony admitted by the court did not show that there was any negligence in consequence of using the rubber gaskets instead of the lead gaskets, and that having taken proper exception to the rulings of the court upon evidence which might have substantiated his position, he is still entitled to insist upon his exceptions. We cannot say that there is nó support for the position of plaintiff’s counsel and that he may not have misunderstood the position of the court in stating his theory of the case. He may have supposed that the court was .stating this theory with reference to the casé as left by the rulings of the court, and, in view of previous admonitions, was not inclined to interrupt. There is also room for the conclusion that his admission quoted above was limited to the case presented under alleged erroneous rulings of the court. We think that testimony which might have been very important was excluded by the court. Plaintiff’s counsel sought to show that prior to the injury to the plaintiff’s intestate certain gaskets on this same high pressure side of the expansion cock had been removed from other joints than the one in question and replaced with others, but the court refused to permit questions with this purpose to be answered, and this forms the basis of the plaintiff’s first assignment of error.
Inasmuch as the gasket is absolutely essential to the tightness of the joint, and inasmuch as the gaskets were all originally put in at the same time in 1898 and the gasket in question had never been inspected from that time until the time of the accident in 1902, the fact that it was found necessary to remove other gaskets and replace them, in joints subject to the same pressure as the joint in question, if it appeared from the answers that such removal was made necessary by the deterioration of such gaskets, would have a very important bearing upon the question whether or not it was the duty of defendant *381to inspect such gaskets from time to time to ascertain their condition.
Plaintiff’s counsel also propounded a question to the witness Heuge, who had testified that they drew off all of the lubricating oil circulating through the • pipes which they could get out in the oil traps, as follows:
“Q. Isn’t it a fact that you cannot draw it all out?
“ This question was objected to, and the court said: ‘ If you cannot draw it out, you cannot help it being there.’
“Mr. Cook: But if I show that it has been there for years—
“The Court: I don’t want .any argument now, Cook. We are talking of the difference of conditions when they draw off all they can through the cock. That is the testimony. They try to have it all drawn off, if possible, through the oil cocks. That is the testimony. Now proceed.”
The further question was put by plaintiff’s counsel:
“Q. What is the effect of oil upon rubber used in making rubber gaskets ?
“A. I think as long as the gasket is tight between the two flanges it couldn’t affect it at all.
“Q. What effect would it have if the gasket was not tight?”
The answer to this question was excluded by the court. It is apparent from some of the testimony which was permitted that oil has a softening effect upon the rubber gaskets, and there was also testimony that the gasket in question was soft and appeared old and rotten. As the tightness of the joint depended to a large extent upon the condition of the gasket, it would seem that, if the effect of the oil necessarily used in the pipes was to soften and disintegrate the rubber, and perhaps make it pervious to the ammonia under its high pressure, the jury might infer from this a duty on the párt of the company to occasionally inspect the gaskets, for the purpose of ascertaining their condition, and this inference would be strengthened if it should appear that other gaskets subject to the same pressure, and under similar circumstances, were found to be defective and removed for that *382reason. It seems apparent that the sense of smell may not be a sufficient notification, as illustrated by this very case, and that even though the joint was tight in the first instance, the deterioration of the gasket, even though the bolts remained in the same position, might eventually render the joint loose. While the court was perhaps justified in understanding the position of counsel for plaintiff, as he stated it, still, in view of the fact that the counsel for plaintiff may not have appreciated the effect of the statement of the court, we are inclined to hold that counsel is not foreclosed from any of his exceptions in consequence of not having called the court’s attention to the inaccuracy of his statement. The duty of inspecting the gaskets is asserted in the declaration; evidence was offered to prove that duty, which was, as we think, erroneously rejected, and it is undisputed that the duty, if found to have existed, was not discharged. We also think that, in view of the relation of witness Heuge to the case and of the apparent necessity imposed on plaintiff of calling him as her witness, plaintiff’s counsel should have been permitted to put leading questions to him and introduce his contradictory statements, if any. Furthermore, we are of the opinion that the presence of oil in the pipes and its effect upon the rubber gaskets was competent. We do not think it can be said, as a matter of law, that, with evidence which the plaintiff might have produced except for the erroneous rulings, the. cause of this accident would appear to be entirely conjectural, or that the case presented would be one devoid of evidence that a previous inspection would have been of value. Warren v. Railway Co., 141 Mich. 298; Schoepper v. Chemical Co., 113 Mich. 582; O’Neill v. James, 138 Mich. 567 (68 L. R. A. 342); McLean v. Railroad Co., 137 Mich. 482; Stowell v. Standard Oil Co., 139 Mich. 18.
The judgment is reversed, and a new trial granted.
Grant, Montgomery, Ostrander, and Hooker, JJ., concurred.