E. T. Bailey brought an action for damages against the Atlanta Gas-Light Company. The jury found for the defendant, and Bailey’s only exception is to the judgment overruling his motion for a new trial. The petition substantially alleges that the defendant damaged the plaintiff in the sum of $10,-000, by installing on the back porch of his home a defective gas meter from which natural gas leaked and entered the adjoining bedroom of the plaintiff through an open window between the porch and the bedroom and so affected his nose, throat, tongue, 'eyes, and bronchial tubes that his throat was inflamed, his sight and respiration impaired, and he suffered “excruciating physical pain and mental agony.” The salient features of the case made by the evidence adduced in behalf of the plaintiff follow: On *79the last day of April, 1934, the plaintiff and his family moved into a house located on Austin Avenue in Atlanta. The defendant had placed its gas meter on the back porch of the house under a large window between the porch and plaintiff’s bedroom. This porch was about fourteen or fifteen feet long and about seven or eight feet wide, and was enclosed by walls on three sides and latticework on the fourth side. "When one of the defendant’s agents came out on April 30, 1934, to turn the gas off for the previous tenant, the plaintiff and his wife were present, and “plaintiff told the gas man that he smelled gas, and the gas man stated it was from changing the meter.” When the gas man came the next day to turn on their gas “plaintiff’s wife told him that she still smelled gas.” Defendant’s agent “tested the joints of all the pipes with soapy water,” and “brushed around the joints,” and said “it was all right,” and plaintiff “dismissed it from his mind, and figured it was probably the odor of creosote” which the former tenant had sprinkled about the place. The plaintiff smelled gas again when one of the defendant’s employees came to read the meter on May 8, 1934. “The night before plaintiff was asphyxiated he noticed no more odor of gas than usual” on the back porch. When he went in and out of his bedroom as usual on May 13, he did not notice the odor of gas. The window between the porch and plaintiff’s bedroom was closed up to May 13, when it was opened because the weather turned warmer. Plaintiff retired at about eight o’clock on the night of May 13, and was awakened at about ten' o’clock when his wife returned from church. At that time he was very drowsy and had a headache. He was “dizzy, and did not notice any odor of gas.” His wife “retired a little while after, with plaintiff in the same room.” The room was closed, except that the window between the porch and the bedroom was open, and there was a slight crack in the door. After the plaintiff went to the door to let his wife in, he went back to bed and “dropped off to sleep.” He waked at about one o’clock and “came to enough to know that something was wrong with him.” He could not speak, and was practically unconscious; but after opening the doors and working with him for some time, his wife got him out of the room. His lungs were congested, his sight impaired, his nose, throat, and bronchial tubes affected, and he could hardly breathe; and though he had been able to work day and night, his ability to work was *80greatly decreased and his earning capacity reduced. He did not call a doctor on the night he was injured, but walked to see two doctors the next day. He continued to go to his place of business after the accident, but could do very little work. On May 14, 1934, one of the defendant’s employees examined the meter and found in it a narrow split about as long as one’s finger.
Dr. C. G. McKay testified for the plaintiff, that he examined the plaintiff in the summer or early fall of 1935, and found that he was “suffering from a bronchial and catarrhal condition of the nose,” and that “his nose was practically filled with proliferation,” which is a growth of the mucous membrane that could be due “to any irritation from smoke chemicals” or cold. Numerous witnesses testified, in effect, that the plaintiff’s health had been excellent before his accident, and poor afterwards. The gist of the evidence introduced by the defendant was that the hole in the gas meter was round and not long, and only large enough to stick a pin in, and was hard to locate, because the ordinary pressure of the gas was not strong enough to make a bubble when soapy water was applied; that the gas was “natural gas” and composed of ninety-seven per cent, methane, or marsh gas, and the remaining three per cent, nitrogen, carbon dioxide and oxjfgen; that this gas is not poisonous and “has no chemical action whatever on the body, or other chemicals, except when you burn it;” that in an ordinary room twenty per cent, of the volume of air is oxygen, but that a person can live when two thirds of the atmosphere is natural gas; that one of the witnesses had worked with this gas for years and had inhaled it in large quantities without any bad effects; that breathing it in large quantities “would have no chemical effect whatever on the nose and lungs,” and could not “possibly do any damage to any tissue of any animal;” that this conclusion was confirmed by extensive experiments on rats; that if there was enough of this gas to crowd out the oxygen a person would die from asphyxiation, but that it would be necessary to have about sixty per cent, of the gas in a room before asphyxiation would result; that if a person were rendered unconscious by an excess of this gas and resuscitated, there would be no chemical effect on the organisms, and he would be in good shape in a short while, except that there might be some effect on the nervous system; and that by actual measurement there were nineteen hundred cubic feet of *81space in the defendant’s bedroom, and it would have been impossible for enough gas to enter the room between the time the plaintiff retired at ten o’clock and one o’clock the following morning to have injured him; and that “if two persons went to bed and one of them woke up at one o’clock in the morning, claiming to be suffering, coughing, and having trouble breathing, while the other did not, that certainly could not be attributed to natural gas.” This evidence was not directly contradicted by any other evidence in the case.
We have made no effort to set out the evidence fully, but have carefully studied the voluminous brief of evidence, and are satisfied that the court did not err in overruling the motion for a new trial. Error is assigned in the first special ground of the motion because the court instructed the jury that “where a plaintiff in a civil case supports his action solely by circumstantial evidence, before he is authorized to have a verdict in his favor the testimony must be such as to reasonably establish the theory relied upon, and to preponderate to that theory rather than to any other reasonable hypothesis.” It is not insisted that the charge was not correct as an abstract statement of law, but it is urged that the “action was supported by positive and direct testimony,” and that the use of the word “solely” generated in the minds of the jurors “the impression that the plaintiff’s contentions were based upon mere imagination . , and a case supported solely by circumstantial evidence would be considered by the jury as a case supported by inferior and light, and probably conjectured, evidence.” The excerpt from the charge is in the language of the first headnote of Georgia Railway & Electric Co. v. Harris, 1 Ga. App. 714 (57 S. E. 1076), is applicable to the evidence in the instant case, and for no reason assigned did the court err in giving it to thé jury. See also Penn Mutual Life Ins. Co. v. Blount, 39 Ga. App. 429, 442 (147 S. E. 768).
Since the excerpt from the court’s charge complained of in the second special ground appears to have been somewhat garbled, we shall state it as it appears in the approved transcribed charge of the court in the record. It is as follows: “Where a plaintiff relies on circumstantial evidence to prove either negligence on the part of the defendant, or that this negligence was the legal and natural cause of the plaintiff’s injuries, where the inference of negligence *82or proximate causation, if existing, depends upon circumstantial evidence, and where there is direct testimony consistent therewith which shows that the defendant was without negligence or that the alleged acts of the defendant could not proximately cause the plaintiff’s injury, and this testimony is not subject to discredit, you should find against the plaintiff on these matters; for a fact can not be established by circumstantial evidence which is perfectly consistent with the direct, uncontradicted, reasonable, and unimpeached testimony that the fact does not exist.” Somewhat generally stated, the assignments of error on this charge are based upon the contention that it was not warranted by the evidence and was contrary to the evidence. The excerpt follows closely the rule stated in Emory University v. Bliss, 35 Ga. App. 752 (134 S. E. 637). The headnote of that case is as follows: “In a suit for damages alleged to have been sustained as a result of the defendant’s negligence, where the inference of negligence, if existing, depends entirely upon circumstantial evidence, and where there is direct testimony consistent therewith which shows that the defendant was without negligence and which is not subject to discredit upon any ground, a verdict in the plaintiff’s favor is unauthorized.” That case involved a gas explosion, and its facts somewhat resemble those of the instant case. The writer of that opinion, quotes as follows from Neill v. Hill, 32 Ga. App. 381 (2 b) (123 S. E. 30) : “A fact can not be established by circumstantial evidence which is perfectly consistent with direct,- uncontradicted, reasonable, and unimpeached testimony that the fact does not exist.” See also Taggart v. Savannah Gas Co., 179 Ga. 181 (175 S. E. 491). Viewing this ground in the light of the issues presented by the pleadings, we are satisfied that the evidence warranted the charge, and we hold that none of the assignments of error thereon warrants a reversal of the judgment.
In special ground 3 error is alleged on the following excerpt from the court’s charge to the jury: “So, in this ease, if you should find that the plaintiff relies on circumstantial evidence to establish the fact that on April 30, 1934, and May 1, 1934, there was a leak in the meter on the premises at 1034 Austin Avenue, but should you also find that the direct testimony introduced by the defendant is consistent with said circumstantial evidence, and that the evidence of the defendant negatives any such fact, then you would *83find for the defendant on that particular charge of negligence.” This charge was authorized by the pleadings and the evidence, and is not objectionable as an alleged expression of opinion by the court “as to the weight which the jury should give to the testimony in the case,” and none of the assignments of error thereon shows cause for reversal of the judgment.
Ground 4 avers that the court erred in defining “asphyxia” and “asphyxiation.” The charge defining these words follows closely the definitions in Webster’s New International Dictionary. The petition alleges that the plaintiff was asphyxiated, and it appears from the evidence that the plaintiff himself used that word in describing his condition. The charge was given in response to a request made by the jury after it had retired to consider the verdict. Error is assigned because the charge “instructed the jury, in effect, that since the plaintiff had alleged he was asphyxiated, that his contention had not been proved unless the evidence had shown that he had been apparently dead, or that his life had been temporarily suspended on account of deficiency of oxygen and excess of carbon dioxide in the blood; whereas movant contends that if he suffered any damage in degree less than apparent death, on account of breathing this gas, . . he was entitled to recover therefor, whether he had been rendered apparently dead or not.” The instruction under consideration was followed thus: “I will instruct you that asphyxiation, in itself and as a concrete definition, does not include poisoning. I think that is the gist of what you want to know, and I will state that asphyxiation in itself does not include poisoning or breaking down or injury to a tissue, but . . . asphyxia is the apparent death or suspended animation in living organisms, due to deficiency of oxygen, and excess of carbon dioxide in the blood, as in interruption of respiration from suffocation or drowning, or from the inhalation of irrespirable gases.” We are satisfied that when the excerpt from the charge is viewed in the light of its context and the charge as a whole, it is not subject to the assignment of error.
Judgment affirmed.
MacIntyre and Guerry, JJ., concur.