The verdict against the company was for $5,000, the amount of the policy, together with over $1,000 for interest thereon, $750 for attorney’s fees, and fifteen per cent, on the aggregate for damages, the total finding being $7,575.27. The court refused to grant a new trial on any of the almost three score and ten grounds embraced in the motion.
The general grounds of 'the motion will not be discussed, save as to one of them, which will be considered briefly in concluding this opinion. The special grounds admit of classification thus: (A) Rulings on the admission of evidence; (B) Rulings on the rejection of evidence; (C) Nature of the plaintiff’s evidence, and its sufficiency to withstand a motion for nonsuit; (D) Conduct of the trial, and irregularities pending its progress.
*758(a) rulings on the admission oe evidence.
1. Was the policy of insurance admissible without the application upon which it was issued? The policy purports on its face to issue “in consideration of the warranties made in the application for this insui’ance, and of the sum of twenty-five dollars”; and it contains a further reference to the application as follows: “Provided, always, that this policy is issued and accepted subject to all the provisions, conditions, limitations and exceptions herein contained or referred to, and upon the express agreement that all the statements and declarations made in the application for this insurance are warranted to be true in all respects, and that said application . . is referred to and made a part of this contract; and if this policy . . has been obtained through misrepresentation, fraud or concealment, . . then the same shall be absolutely void.” The last condition in the policy declares that “all the provisions and conditions aforesaid, and a strict compliance therewith during the continuance of this policy, are conditions precedent to the making of this contract.”'
The action was complaint, and the plaintiff’s petition or declaration said nothing of the application; it pleaded the policy only, and set out a copy, the copy being annexed to the petition. No objection was made ' to the petition by demurrer or otherwise. The defendant treated the plaintiff’s pleading as sufficient, and the evidence offered coincided therewith; the document counted on, described and copied was the one offered and admitted. Moreover, tested by the code, §3392, the pleading was quite sufficient; and we think it follows that in making a prima, facie case for recovery, the actionds to be treated as founded on so much of the contract as is set forth in the policy, leaving stipulations, warranties and conditions expressed only in the appli- . cation to be brought to the notice of the court defen*759sively by the company. This view of the relation of the policy to the application, is perhaps sound independently of statutory provisions. We are inclined to think that, upon general principles of modern pleading and practice, a policy of this kind can be pleaded and proved separate and apart from the application, though authorities on the subject differ. The affirmative is more or less supported by Throop v. Ins. Co., 19 Mich. 424; Ins. Co. v. McCookey, 33 Ohio St. 555; Redman v. Ins. Co., 49 Wis. 431; Cont. L. I. Co. v. Rogers, 119 Ill. 475; Guardian M. L. I. Co. v. Hogan, 80 lb. 35; M. B. Ins. Co. v. Robertson, 59 lb. 123; Herron v. Ins. Co., 28 Ib. 235; Simmons v. Ins. Co., 8 W. Va. 486; Roach v. Mutual Co., 28 S. C. 431, 6 S. E. Rep. 286; Cont. Ins. Co. v. Kessler, 84 Ind. 310; Penn Ins. Co. v. Wiler, 100 Ib. 92. And see Mutual Co. v. Cannon, 48 Ind. 264; Com. Ins. Co. v. Monninger, 18 Ind. 352; Piedmont Ins. Co. v. Ewing, 92 U. S. 377. Contra, Bobbitt v. Ins. Co., 66 N. C. 70, 8 Am. Rep. 494; Bidwell v. Ins. Co., 3 Sawy. 261; Gilmore v. Ins. Co., 55 Cal. 123; Rogers v. Ins. Co., 72 Iowa, 448, 34 N. W. Rep. 202; Lycoming Ins. Co. v. Sailer, 67 Pa. St. 108; Am. U. Ass. v. George, 97 Ib. 238. And see Bacon on Ben. Soc. & Life Ins. 463; Bliss Life Ins. §61, p. 91; 2 Wood Eire Ins. 1119, 1136.
2. Was it error to admit in evidence the affidavits which had been produced to the company as preliminary proof of injury and death? These affidavits were made by Boykin, Turner and Brown respectively. The contract of the company as expressed in the policy was to pay “within ninety days after sufficient proof that the insured, within any time during the continuance of this policy, shall have sustained bodily injuries effected through external, violent and accidental means, within thé intent and meaning of this contract and the conditions thereunto annexed, and such injuries alone shall *760have occasioned death within ninety days after the happening thereof.” The declaration alleged that “proper and sufficient proof” had been furnished of accidental death, and that payment had been demanded. The company, by plea, traversed the whole declaration, averring that the allegations therein were not true, and that at the time of bringing the action, Sheppard was not dead but was then in life.
The policy, under the head of conditions, contains a further clause, as follows:
“ In the event of any accidental injury for which claim may be made under this policy, immediate notice shall be given in writing, addressed to the secretary of the company, at Hartford, Conn., stating the full name, occupation, and address of the insured, with full particulars of the accident and injury; and also in case of death resulting from such injury, immediate notice shall be given in like manner; and failure to give such immediate written notices shall invalidate all claims under tMs policy; and unless direct and affirmative proof of the same, and of the death, or duration of total disability, shall be furnished to the company within seven months from the happening of such accident, then all claims accruing under this policy shall be waived and forfeited to the company.”
There was no suggestion by counsel for the defendant that the affidavits tendered in evidence were not produced to the company in due time, as preliminary proof of injury and-death. The objections stated and urged were, that the affidavits were hearsay, irrelevant, and not being positive and direct proof they were not the character of proof required by the policy; also, that they were not admissible upon the question of demand, because the company, by refusing to pay on the ground that Sheppard was alive, waived demand, and thus proof of demand was unnecessary. The court, after receiving the evidence, instructed the jury that the affidavits were admitted, not to prove the fact of death, but for the *761purpose of showing that the proofs were submitted to the company within the time required by the policy.
If by affirming that demand had been waived and saying that proof of demand was unnecessary, the counsel meant that the company had waived, or did then waive, compliance with the condition in the policy requiring preliminary proof of injury and death, the counsel should have so announced in plain terms. In the state of the pleadings, nothing short of an explicit waiver would entitle him to exclude the evidence as irrelevant or unnecessary. By the pleadings, the parties were at issue no less upon the question whether preliminary proof had been furnished according to the condition of the policy, than upon the ultimate questions of accident and death. It was not to prove a demand for payment that the affidavits were offered, but to show that a condition precedent to a right of action on the policy had been performed. No doubt, performance of the condition, or proof of its performance, could be waived at the trial, and if done expressly, especially if the waiver were incorporated in the pleadings, which would be easy by amending the answer, any evidence on the subject would be needless and therefore irrelevant. Such evidence has but one legitimate object, and that is to meet the condition of the policy as a preliminary to bringing action where voluntary payment has been denied. In supporting an action, beyond clearing the way for it, the preliminary proof has no efficacy whatever. So. Ins. Co. v. Lewis, 42 Ga. 587; Mutual Co. v. Stibble, 46 Md. 302; Ins. Co. v. Gould, 80 Ill. 388; Newmark v. Ins. Co., 30 Mo. 160; Breckinridge v. Ins. Co., 87 Mo. 63; Sexton v. Ins. Co., 9 Barb. 191; Howard v. Ins. Co., 4 Denio, 502.
Indeed, the general and better opinion seems to be that, where the preliminary proofs are in writing, the only concern of the jury with them is to pass on their *762authenticity and identify them as the proofs submitted to the company. The question of their sufficiency is for the court. Klein v. Ins. Co., 13 Penn. St. 247; Com. Ins. Co. v. Sennett, 41 Penn. St. 162; Kittanning Ins. Co. v. O’Neil (Pa.), 1 Atl. Rep. 592; Citizens Ins. Co. v. Doll, 35 Md. 89; Mutual Ins. Co. v. Stibble, 46 Ib. 302 (supra); Neese v. Ins. Co., 55 Iowa, 604; Gooche v. Ins. Co., 4 Woods, 102, 10 Fed. Rep. 347.
3. Yfas the objection well-founded, that the affidavits were hearsay, and not the character of proof required by the policy, because not positive and direct proof? As this objection was not confined to any particular affidavit, but was urged generally against all, it is enough to set out here one of the affidavits and consider the question -with reference to it alone. The affidavit of Boykin was sworn to on the 19th day of February, 1885, and the statement contained therein was as follows:
“My name is Kedden Boykin. I live and farm near Bristol, which is my post-office address, in Liberty county, Florida. -I knew Thomas J. Sheppard and was very well acquainted with him. On the 5th day of January, A. D.,1885,1 was with Thomas J. Sheppard, Felix Brown and Alexander Turner.- On the day above mentioned, we went together on a hunt; we went in a common bateau through the river swamps to the Appalachicola river. We took or got into a bateau at Sheppard’s orange grove; we reached the river, I suppose, at about ten o’clock in the forenoon. We crossed over the river and hunted awhile, then the party came back, and took the boat for a point above on the opposite side of the river. We left Felix Brown, when Turner, Sheppard and I took the boat, to take care of some dogs until we returned. We went up the river to a point about half a mile from where we left Brown. Turner and myself got ashore to drive or hunt deer down the river. Sheppard remained in the boat and was to take the boat down to a point where Turner and I were to join him and take the boat for the camps *763at Sheppard’s grove. Turner and I separated, hut went about in the same direction; I was sometimes fifty, sometimes one hundred yards or more from the bank of the river; I was coming in or near the bank when the accident occurred. I heard the report of the gun and looked at once in the direction of the gunshot. I glimpsed Sheppard falling; I distinctly heard him strike the water. There was some woods or brush between me and Sheppard; I was, I think, about forty yards away; I went quickly to the place where I found the boat drifting several yards from the hank containing the hat and gun of Sheppard. I called at once for Turner to come there. Turner was some farther away than I. When Turner came, we cut poles and sounded for the body of Sheppard. A short distance from the bank we found the water too deep and the current too strong; we could not do much in that way. The water a few yards from the bank was about twenty feet deep. The accident occurred about four o’clock in the afternoon. Turner and I took the boat and went down the river to where we left Brown, and took him aboard and went to the camp at the grove, which we did not reach till after dark. To the best of my knowledge, recollection and belief, Sheppard shot himself in attempting to walk ashore from the stern of the boat.”
In this statement there is no hearsay, hut it is true that injury and death are not established by “direct and affirmative proof,” without the aid of inference. It was probably the design of the company in framing the policy, to exact preliminary proof in which inference should have no share, but if so, countenance cannot be given by the courts to such an attempt to vary the ordinary rules of evidence. We shall see hereafter that this cannot be done as to the ultimate proof at the trial of an action founded on the policy, and it would certainly be inconsistent to uphold a stipulation for higher or more complete proof as a condition to a right of action, than would be requisite to support the action itself. To do this would be like upholding a condition in a grant repugnant to the grant and allowing the *764main object and purpose to be defeated by a condition. If the terms “direct and affirmative proof/’ as used in the clause of the policy we are considering, admit of more than one construction, they should receive a construction which will harmonize them with the rules of law on the subject of evidence; and if they admit of one construction only, and that is at variance with these rules, the clause must give way to them in so far as the conflict extends. Preliminary proof, satisfactory, or which ought to be satisfactory to establish loss, is all that can be reasonably exacted; and we think the affidavit of Boykin can well be treated as a sufficient compliance with the policy on that subject, in so far as the requirement is compatible with law. By the aid of inference, both accident and death might be deduced from the facts testified to directly in the affidavit, if no rebutting or explanatory circumstances appeared.
4. Was the affidavit of W. B. Sheppard, a brother of the assured, admissible in evidence to chai’ge the company with notice of the information therein contained, for the purpose of illustrating the good or bad faith of the company in refusing payment ? This affidavit was sworn to in J uly, within the seven months limited by the policy for producing preliminary proof. Its contents consist of a recital of the circumstances attending the disappearance of Sheppard, based, no doubt, mainly on hearsay, together with some representations as to a reward having been offered for the recovery of his body, and as to the character of the river and the nature of the obstacles to recovering the body of a. person drowned therein. It furnished the names and address of numerous citizens as persons acquainted with the river and having knowledge of these obstacles. Perhaps there might be no valid reason against treating this affidavit as supplemental- to the proofs of death previously submitted, and as combining with them to *765strengthen their force and effect as preliminary verification, but it was not competent evidence for the purpose of subjecting this company to the imputation and the consequences of bad faith. The faith of the company should not be judged by the preliminary proofs or other ex parte affidavits, but by the case made at the trial. The preliminary proofs go to the liability, not to the faith of the company. Its duty to pay, without delay or resistance, would arise out of the fact of accidental death and the reception of sufficient preliminary proofs, and that duty could not be made more obligatory by any additional information volunteered by the plaintiff before or after suit. In refusing payment after due demand according to the statute, the company would act at its peril, a peril neither increased nor diminished by the amount of information it might have or obtain, hut only by the weakness or strength of its defence as manifested at the trial, any weakness in the plaintiff’s case being, of course, counted as part of the strength of the defence. A defence going far enough to show reasonable and probable cause for making it, would vindicate the good faith of the company as effectually as would a complete defence to the action. On the other hand, any defence not manifesting such reasonable and probable cause, would expose the company to the imputation of bad faith and to the assessment of damages therefor under section 2850 of the code. It would be a dangerous practice to let one party pelt the other with ex parte affidavits at will, and then bring these affidavits in at the trial to raise or support the charge of bad faith. This affidavit is no legitimate evidence for the jury of any fact which it recites or affirms. Lycoming Ins. Co. v. Schreffler, 44 Penn. St. 269. Nor would it be, had it been made by agents or employees of the company. Carroll v. R. R. Co., 82 Ga. 452. It is mere “side-wind” evidence, a *766class of evidence than which none not absolutely false can be more pernicious. Chief Justice Tilghman, in Thurston v. Murray, 3 Binn. 328, in which case a paper nad been read to the jury, not as evidence of the truth of the matter contained in it, but as evidence of a communication from the assured to the underwriter, said: “This point involves a question of considerable importance, whether the assured shall be permitted to bring before the jury papers which in themselves are not legal evidence. It is agreed that the copy of the decree was not per se evidence. Why then should it be read? Because, say the plaintiffs, it was exhibited to the defendant’s broker as one of the proofs of loss, and it is expected by the underwriters that these proofs of loss should always be exhibited. Upon the same principle, every paper which the assured wishes to read, however improper in itself, may be brought before the jury ; for it is in his power to lay before the broker what papers he pleases. . It appears to me that it will be of ‘dangerous consequence if papers are thus introduced by a side-wind, which not being evidence in themselves could not be brought forward directly.”
5. The plaintiff, Mrs. Sheppard, testified in her own behalf^ and was thus before the jury both as party and witness. After the evidence for the defence was all in, numei’ous witnesses were introduced by the piaintifl‘, who testified that her personal character for integrity, honesty and uprightness was eotLsidered good — first-class ; that they never heard anything against her, and would believe her as soon as they would anybody. Was this testimony admissible? No attack upon her character had been made either by general evidence to impeach her credit, or by .proof of previous statements inconsistent with her testimony as given at the trial. Thus, in the capacity of witness her character was not in issue. Was it in issue in the capacity of party to the *767suit ? The suit was for an alleged breach of contract. The defendant denied any breach, and set up that the life insured had not terminated, but was still in existence. Certainly the nature of the .action did not involve the general character of the plaintiff, nor was there any imputation derogatory thereto, express or implied, in defending upon the general issue or the special plea. Thus, whether we look to the suit as a whole or to either side separately, the nature of the controversy does not involve the character of either of the parties. Neither by the pleadings nor by the evidence does the defendant impute to the plaintiff any complicity in the disappearance of her husband, or any personal knowledge that he is still alive. And her honesty and good faith in prosecuting the action can-add nothing to its strength, nor would her bad faith, were she chargeable with any, aggravate its inherent weakness. If her husband be in fact dead, and it be so made to appear by the evidence, she ought to recover and would l’ecover, although she might believe him still in life. Her honesty or dishonesty, her rectitude or the want of it, has nothing whatever to do with the merits of her claim, or the measure of her recovery. Whether she is a good woman or a bad one, her rights as a suitor in this litigation are the same ; and her own merits or demerits as a lady can throw no light, according to the principles of forensic evidence, upon the question whether her husband is dead or alive, the one decisive question upon which the result of the suit depends. The code (§3757) and the common law coincide in restricting the right of a party to adduce evidence of his own general character to cases in which the nature of the action involves such character. Nash v. Gilkeson, 5 S. & R. 353; Anderson v. Long, 10 Ib. 55; Atkinson v. Graham, 5 Watts, 411; 5 Smets v. Plunket, 1 Strob. 372; Fowler v. Ins. Co., 6 Cow. 673; Gough v. St. John, 16 Wend. 645; Givens *768v. Bradley, 3 Bibb, 192; Potter v. Webb, 6 Me. 6; Gutzwiller v. Lackman, 23 Mo. 168; Dudley v. McCluer, 65 Ib. 241; Church v. Drummond, 7 Porter (Ind.), 17; Bartow v. Thompson, 56 Iowa, 571, 41 Am. Rep. 119 and notes; DuBose v. DuBose, 75 Ga. 753. Courts differ, however, sometimes in applying the rule. Contrast Henry v. Brown, 2 Heisk. 213, with some of the above. Also, Spears v. Ins. Co., 1 Baxter, 370, and Mosley v. Ins. Co., 55 Vt. 142, with Schmidt v. Ins. Co., 1 Gray, 529, and Continental Ins. Co. v. Jachnichen (Ind.), 10 N. E. Rep. 636. Also, McNabb v. Lockhart, 18 Ga. 495, and Falkner v. Behr, 75 Ga. 672, with Morris v. Hazelwood, 1 Bush, 208, and Wright v. McKee, 37 Vt. 161. But no court, so far as we know, has ever held that a case like the present is one for the admission of evidence of general character for the vindication of either party. The honor of neither party would be tarnished by the result of the suit, no matter what the result might be. No doubt a verdict rendered on the basis that Sheppard is alive would tend to disgrace him, but it could have no just bearing against the personal honor or reputation of his wife. And what is more to the purpose, her fair repute and good name can have no probative force whatever upon the question of whether he is alive or dead. The case of McNabb v. Lockhart, 18 Ga. 495 (supra), was urged upon us by counsel in a supplemental brief devoted to it alone. That case is certainly not in point, for McNabb set up that he had lost the money which had been entrusted to him as a bailee, and to recover which the action was brought, and whether he had lost it or stolen it was the ultimate question involved. On that question his good character might throw some light, and a decision against his defence would affix a stain upon his reputation for honesty.
6. Boykin and Turner were the persons engaged with Sheppard in the hunt for deer which terminated in *769Sheppard’s disappearance. Boykin died before' the trial, and no testimony from him, other than his affidavit taken as preliminary proof, was introduced. Turner was examined by interrogatories, and a portion of his testimony, some of it in answer to direct and some to cross-interrogatories, was, in substance, as follows: “Boykin and myself were to go down the river by land, and Sheppard was to run down in a boat, about three quarters of a mile, to a certain shanty, there get out and make a stand. We were to rejoin him at that point. I went out 100 or 150 yards from the river, and Boykin was between the river and me, but how far from the river I do not know. As we were driving along down the river for deer, I heard a gun fire in the direction of the river, and afterwards some noise, hut for some time did not stop to listen. When I did stop I heard some one calling me excitedly, and X went in that direction, and in a very short time met Boykin coming towards me in a run, and calling as he came. IXe was very much excited and looked very wild. When he came up he said, Tom Sheppard had killed himself, or had shot himself; which of these expressions he used I do not remember. We hurried hack to the place. On the way back he said he heard the gun fire and heard a splosh in the water, and thought he glimpsed Sheppard as he fell from the boat. He said he was right out there, indicating with his hand a spot about 15 or 20 yards from where Sheppard was, and there was a thick strip of cane-brake between him and Sheppard. He said he heard a splash in the water as of some object striking it. After we reached the river-bank I shot off my gun, then got in the boat and examined it. It was fastened, and was floating on the water about three feet from the bank. There was no blood on it, and I could see none on the water. Sheppard’s gun and hat were in the boat, the hat just in front of the middle seat, and the gun lying up and down the *770boat in real- of that seat, with the butt against the middle piece. It was a breech-loading shot-gun, with hammers known as rebounding, without notches in the tumbler to hold at half-cock, the hammer being held above the tube by a spring. One of the hammers was ofi; only one barrel would fire. It was empty except a shell. With long poles Boykin and I made search in the river for Sheppard’s body, but could not find it., After a short time, the search was discontinued because night was approaching. The boat was found about 200 yards below where Boykin and I got out, and about 100 or 150 yards from where I Avas Avhen the gun fired, at which time I had gone about 300 or 400 yards. I saw Boykin in a feAV minutes after I heard the gun.” Were the declarations made by Boykin to Turner admissible evidence as part of the res gestee ? The fact in issue was the accidental death of Sheppard. No witness saw him die or knew certainly that he was dead. Ilis death, if established at all, could be established by circumstantial evidence only. Amongst the most direct and material circumstances bearing upon the question of his death, was his disappearance as detailed in, the evidence of Turner above quoted. The disappear- | anee as a Avhole was a material evidentiary fact relevant to the matter in issue. As a whole it was a composite or complex fact consisting of divers particulars, including the hunt, its abandonment by Boykin and Turner, and the search which immediately ensued. To understand the significance of the disappearance it is necessary to investigate both the hunt and the search, together with the circumstances attending them. The hunt was a concerted enterprise for joint execution upon a given scene by three actors, each of whom assumed a part and entered upon its performance. It was never finished, but was abandoned first by Boykin, then by Turner. There can be no doubt that the abandonment *771of the hunt aud institution of the search constituted a part of the res gestee of the disappearance. For, though the disappearance as a physical fact was complete before the search began, as an evidentiary fact to throw all the light that it is capable of throwing on the question of death, it was incomplete until it was verified by the results of the search. Boykin and Turner, as well as Sheppard, were actors in the disappearance, and their acts as throwing light upon it are to be investigated and explained. The disappearance did not consist alone of Sheppard’s exit from the boat, but included his subsequent absence therefrom and failure to return, until a reasonable time had elapsed for his return had his absence been temporary only. It embraced both affirmative and negative elements, and, as a whole, was a continuous act. Its time relation was unchanged until, failing to return, he was looked for and could not be found. It is certain that Boykin did not abandon the hunt and enter upon the search by reason of the same causes which influenced Turner. Turner heard the report of the gun and at the same time some noise, but these alone did not induce him to give up the hunt. He remained in it until he heard some one calling him, and then did not enter upon the search until after he had met Boykin and heard a part of his declarations. Without knowing what Boykin said to him when they met, it would only be matter of conjecture why Turner abandoned the hunt permanently and engaged in the search. And without a recital of the subsequent declarations, those made whilst he and Boykin were on the way to the river, it would be matter of conjecture only as to'what induced Boykin to leave the hunt, seek Turner and engage in the search. The search itself being a relevant and material fact, declarations made at the time calculated to account for and explain it are also relevant. The transaction which was in-.progress *772when tlie disappearance occurred was the hunt in which Boykin and Turner, as well as Sheppard, participated. The hunt passed, without any interval between them, into the search; the same actors save Sheppard were in both; and the stream of action, though it swerved abruptly from its original course, flowed in an unbroken current till the search was concluded. Morever, the scene of action was unchanged; as compared with that • in contemplation when the hunt began, it was contracted, but was not otherwise varied. There were acts done by Boykin which were meaningless without some verbal explanation. He had abandoned the hunt, was calling for Turner and was running towards him. Why? He explained why by his declarations. They were needed then to illustrate his conduct; they were also needed to illustrate the conduct of Turner, who, instead of going back and resuming the hunt which he had left temporarily on hearing some one calling him, abandoned the hunt, 'went with Boykin to the river, examined the boat and searched in the water. These acts of Turner would not be explained by the conduct alone of Boykin, but are explained by that and the declarations taken together. Moreover, so far as he was concerned, the declarations doubtless had an influence on the scope and manner of the search. Why did he direct his examination to the boat and the water alone, hot extending it to the land? So far as appears, he did not even look for tracks on shore. As the boat was tied when-he reached it, this would be an unaccountable omission, were it not for the light thrown on it by what Boykin had said as to his having caught a glimpse of Sheppard and heard him or some object strike the water. With the information which Boykin professed to have, it was his duty to communicate it to Turner, and the duty of both to act upon it as they did. To make the communication would be the natural and spontaneous *773impulse of any man under like circumstances. The mere human machine, run by vital and emotional forces alone, with no conscious exertion of the will, would prompt such a disclosure in almost every conceivable case under like circumstances. The declarations must have been made very -shortly after the gun fired; for it was only a few minutes until Turner saw Boykin, who was then running towards him. Some of the declarations were made as soon as they met, and the rest whilst they were hurrying back to the river. The distance they had to go is not stated. But it could not have been far, as Turner was only 100 or 150 yards from the boat when he heard the report of the gun, and from that time till he saw Boykin was but a few minutes. Taking all the evidence together, there are indications that Boykin had been to the river, and finding the boat drifting, had hauled it in and tied it to the bank before he started in quest of Turner. But this could have occupied only a short time, and was not such an intervening action on the part of Boykin as to separate his subsequent proceedings and the declarations accompanying them from the antecedent stages of the general transaction. -It would be altogether natural, if Boykin saw and heard what he reported, for him to go first to the boat, and having secured that, to hurry instantly to Turner with a view of procuring his assistance. Doubtless all he did was in hot haste, and with no pause or interval in the rush of his movements. The impression made upon the mind by Tuimer’s evidence is that only. a brief time — a very brief time— elapsed after he heard the report of the gun until he appeared with Boykin bn the bank of the river where the boat was found. lie speaks of a few minutes, of running and of hurrying, from which it is natural to infer rapid action and brief time. It cannot be correct to treat Boykin and Turner as mere spectators in the *774scene of Sheppard’s disappearance. They were not merely in the environment of the event; they were a part of its organism; actors in it, parties to it, if not on its active, certainly on its passive side. Sheppard did not merely disappear, he disappeared from them. It was upon them that his disappearance smote, as it were, directly. Upon them the blow fell, and their conduct in receiving it illustrates its character and the manner in which it was delivered. On the principle of action and reaction, its immediate effect upon them tends to illustrate its nature and character. This being so, the declarations, made by one and heard by the other, and the matter of them influencing both in their action and conduct, were pertinent and material. There were acts done by them which constitute good primary evidence in this case, and the declarations throw light upon and explain them.
But looking at the declarations in their relation to Sheppard’s acts alone, and regarding Boykin and Turner in no wise as actors themselves, further than that they were engaged at the time in participating with Sheppard in the hunt, it would be no strain to treat the declarations as a part of the res gestee from that standpoint. The departure of Sheppard in the boat for a point three quarters of a mile distant, the finding of the boat two hundred yards from the place of departure, with Sheppard no longer on board, would indicate with fair certainty that he left the boat at some point within the two hundred yards’ distance. When, where and how he left it are matters of inference from all the circumstances taken together. Some of these circumstances are the following: Boykin’s route lay nearer the river than Turner’s. Turner, from the point of departure, which was the same for all three of the hunters, had gone three or four hundred yards when he heard the report of a gun towards the river. He went *775but little further until he heard some one calling him, and presently he saw Boykin running towards him. They met; Boykin exclaimed that Sheppard had shot or killed himself. They htirried- forward to see about it. On the way, Boykin explained his reasons for thinking what he had just announced. They arrived at the river, found the boat, Sheppard’s hat and gun, but Sheppard was missing. Of three comrades in a pending hunt one was gone, and one of the two left had made a statement to the other as to what he heard and saw in regard to the matter. The three had a common interest in the hunt, and the two left, a common interest in the fate of the one who was absent. The hunt had been suddenly and unexpectedly cut short by this startling episode, and Boykin was the first to discover and report it, and his report was made within a few minutes after the gun fired. These facts are all vouched by sworn evidence. Were Boykin’s declarations so related to the main fact, of which fact, to wit, Sheppard’s separation from the boat, there can be no possible doubt, as to be a part of the res gestee of the same? “ Declarations accompanying an act, or so nearly connected therewith in time as to be free from all suspicion of device or afterthought, are admissible in evidence as part of resgestee.” Code, §3733. What the law altogether distrusts is not after-speech but afterthought. The code introduces no new rule, but frankly recognizes in its letter the full breadth of the temporal element in the rule which it found existing, as expounded in the luminous and able opinion of Judge Nisbet in Mitchum v. State, 11 Ga. 615, an opinion delivered in 1852, and so excellent in thought and expression as to tempt a New Hampshire judge, in Tucker v. Henniker, 41 N. H. 317, to assume to himself the authorship of a considerable portion of it relating to another subject. The rule contemplates that all the res *776gestae, including declarations forming part thereof, must transpire within the present time of the transaction. But that time, while it cannot be less, may be more extended than the present of the principal fact, in some instances a little, in others much, and in others very much more. Usually if they can all be ascertained, some of the res gestae will be found simultaneous with, and some anterior and others posterior to the principal fact. Thus, suppose an electric discharge during a summer shower to be the principal fact, the formation of the cloud, the falling of the rain, the- thunder and its reverberation would all, for some purposes, be within the res gestae of the event, though the principal fact was but a flash of lightning. This example may serve as a figure to chai’acterize the instances in which declarations subsequent to the fact are receivable in evidence. Let thunder represent mental impressions produced by the event. Then reverberation will represent admissible declarations reporting these impressions. It will represent them, by a close analogy in two respects, first in being speedy, second in being spontaneous. That they shall be or appear to be spontaneous is indispensable, and it is for this reason alone that they are required to be speedy. There must be no fair opportunity for the will of the speaker to mould or modify them. His will must have become' and remained dormant, so far as any deliberation in concocting matter for speech or selecting words is concerned. Moreover, his speech, besides bbing in the present time of the transaction, must be in the presence of it in respect to space. He’ must be on or near the scene of action or of some material part of the action. His declarations must be the utterance of human nature, of the genus homo, rather than of the individual. Only an oath can guarantee individual veracity. But spontaneous impulse may be a sufficient sanction for the speech of man as such, man as distin*777guished from this or that particular man. ' True, the verbal deliverance in each instance is that of an individual person. But if the state of his mind be such that his individuality is for the time being suppressed and silenced, so that he utters the voice of humanity rather than of himself, what he says is regarded by the law as in some degree trustworthy. Boykin’s connection with Sheppard, his interest in what befell him and his relation to Turner rendered it proper for him to make the communication which he did to Turner, and he made it upon the first opportunity, which opportunity was gained by running in quest of him, and in such a short time as reasonably to exclude any suspicion of device or afterthought. Not death only, but accidental death must appear in order to support the action. As bearing upon the element of accident, any evidence is relevant which tends to show that Boykin did not kill Sheppard. His declarations throw light upon that question. Were he charged with the homicide and on trial for it, his conduct and declarations so soon after the gun fired would be competent evidence in his behalf. Suppose, instead of saying to Turner what he did, he had told him that he himself shot Sheppard and threw him into the river, would not this have been evidence for the company, as part of the res gestae, to disprove accidental death ? While we think it is not necessary to invoke the rule of discretion, yet under the operation of that rule it is safe to hold that there was no error in admitting the evidence.
7. Considei’ed in the light of the circumstances, it is obvious that the exclamation that Sheppard had killed himself or had shot himself, was matter of inference or opinion on the part of Boykin. Such being its nature, it would be proper to instruct the jury that this would be evidence for no purpose but to explain the conduct of Boykin and Turner. Were Boykin himself alive and *778on the stand as a witness, lie would not be permitted to express to the jury his opinion, even though he superadded the reasons on which it was founded. For the question whether Sheppard shot himself or not is matter of fact to be found by the jury, and not matter of opinion appealing to the skill or sagacity of a witness. There being no objection to the charge of the court, we are to presume that, on this subject, as well as every other, it was correct. And to this part of Boykin’s declarations no objection was taken on the ground that they involved inference or opinion. To the latter part of the declarations, however, this objection was taken. "We suppose it was intended to apply to the expression that he, Boykin, thought he glimpsed Sheppard as he fell from the boat, this being the only seeming reference to opinion which we find in the whole statement outside of the exclamation which we have just considered. Our construction of the expression is, that it embodies no inference but a direct reference to an immediate perception of the senses. Such perceptions may be more or less dim or indistinct, and by the use of the word “thought,” Boykin meant to signify that his ocular perception of something falling from the boat was indistinct, but nevertheless that*' he had such a perception and considered it reliable.
8. Boykin’s conduct in calling for Turner, running in search of him, etc., being a part of the res gestee, his manner and appearance when they met were also relevant. That he was very much excited and looked very Avild was competent evidence, and was not amenable to the objection that it was mere opinion without any disclosure of the reasons or facts upon Avhieh it Avas founded. An observer may testify to the exhibition by another of excitement or any emotion such as alarm, Avithout specifying the various minute facts which indicated the same to the mind of the witness. The *779emotions register themselves in the appearance and manner, and when a witness testifies to the existence of the emotion, his plain meaning is that he observed the numerous and often nameless indications by which that species of emotion is commonly manifested.
9. The fair and natural limitation of the res gestee of the disappearance is the close of the search made by Boykin and Turner upon the scene of the disappearance itself. This shuts out certain testimony of Brown, who, though a member of the general hunting expedition, had been left behind some distance down the river, when the last hunt was entered upon, to take care of the dogs; this testimony being that, when Boykin and Turner returned in the boat to where he was, they appeared to be very much excited, but Turner was more excited than Boykin. It also shuts out the evidence of Turner, in which, speaking of what occurred at the camp that night, he said he was awake some time in the forepart of the night and they talked about the matter until he went to sleep. Also the evidence of Brown, that as long as he was awake, they talked about the accident that had occurred that day, that their conversation was of a distressed kind, and that as long as they were up, they were discussing the disappearance of Sheppard. Whilst this evidence is not of a nature, perhaps, to harm the case in any way, yet it is wanting in relevancy to the issue, and in our opinion ought, for that reason to have been excluded.
10. The plaintifi proved by a brother and a brother-in-law of Sheppard that the family of Sheppard “recognized him as being dead”; by another brother, that “the family regard him as dead”; and by her own brother who lives in the family with her, that “ the family recognize him as dead.” Was this evidence admissible ? The question in issue was not simply as to death, but death by accident within the year covered *780by the policy, which was from June, 1884, to June, 1885. The belief which sprang up in the family must have taken its rise from particular facts, all of which were so recent when this suit was brought in January, 1886, that they were likely to be susceptible of proof at the trial. They were the premises from which any conclusion by members of the family must have been drawn; and as these same premises, together with othei’S less influential in the family, could be placed before the jury, there would be no necessity to resort to family opinion or belief as evidence to aid the jury in reaching a right conclusion for themselves. The^admissibility of family reputation rests on necessity. Code, §8770. And where no necessity has arisen from lapse of time or otherwise, it is much safer to confine the evidence of death to the original facts and exclude the derivative family opinion as irrelevant. Though some members of the family may have died before the trial, enough of them survived to acquaint the jury with any facts known to the family tending to establish the death of Sheppard; and his wife, two brothers and a brother-in-law, besides her own brother, were examined as witnesses. Moreover, any opinion which could have been formed in the family had to take its rise in company with an interest on the part of Mrs. Sheppard in the subject-matter of this action and in some other policies of insurance on the life of her husband. In a case where the fact of death is directly in question and where the whole pressure is upon that fact together with the manner of its occurrence, no element of pedigree being involved, and the death, if any, being so recent that scarcely one year had elapsed when the action was brought, family belief or opinion would seem as likely to mislead as to lead correctly, especially if the principal member of the family as now existing would have a strong motive of interest to mould the family *781opinion in one way rather than in another. Our code, §3772, recognizes family reputation as a means of proving pedigree, including the fact of death with other facts embraced in genealogical inquiries. Here pedigree is not in question, and the whole spirit of the law of evidence is adverse to allowing the plaintiff to support her action, under the circumstances, by proving her own belief and that of other members of her family as tending to establish the real existence of the principal fact involved in her cause of action.
11. As there was no imputation upon Mrs. Sheppard of complicity in any scheme of simulated death, her grief and sickness consequent upon hearing of her husband’s death or disappearance were irrelevant. So too was her destitution and the sale of clothing to supply her wants or those of her child. The evidence as to these matters was improperly admitted.
12. The plaintiff testified that the relations between herself and husband were pleasant; that if she had not loved him she would not have married him, “which continued up to the time of his death.” Mrs. Ivery, who was an inmate of the family when the disappearance occurred, and had been for some three months previously, testified that “ she never heard any quarrel between plaintiff and her husband; they were kind to each other; certainly there was love between them; saw nothing to indicate anything to the contrary.” This evidence was admissible as throwing light upon Sheppard’s prolonged absence and the probable cause of it. With happy domestic relations there would, according to the ordinary motives of human nature, be less cause for such an-absence, were it dependent on his own will, than if those relations had been different. In this way the fact that Sheppard and wife lived on affectionate terms would tend somewhat to negative any theory either of suicide or of voluntary desertion of *782his home and family. Certainly a husband whose affections were in a normal state and fully reciprocated, would be more likely to return or be heard from were he not dead, than he would were he under the influence of different sentiments at the time of his departure. Authority, as well as the reason of the thing, favors the admission of the evidence. Tisdale v. Ins. Co., 26 Iowa, 170, 176.
13. Miss Horton, plaintiff’s sister, testified that she, the witness, had nothing to lead her to believe that Sheppard was alive. This was objected to as irrelevant. The objection was well-taken. The belief of the witness being irrelevant, her having seen nothing to induce her to form an opinion stands upon the same footing. She was competent to testify to any relevant matters of which she had knowledge, whether her knowledge went to the affirmative or the negative of such matters. But their sufficiency to generate in her a belief would have nothing to do with the case. The motion for a new trial (37th ground) imputes to her a further statement as follows : “ I have never seen any communication, or letters in a mysterious way, to lead me to suppose that my sister, Mrs. Sheppard, was in communication with T. J. Sheppard.” In the brief of evidence this statement is not found. The brief represents her as saying: “Have nothing to lead to believe my sister has been in communication with him.” In either of these forms we think the evidence would be incompetent, for the same reason which we have just pointed out respecting her belief as to the death of Sheppard. To say that nothing is known to cause belief, and that no belief exists, are virtually one and the same thing. To disclaim knowledge which would induce us to believe is, therefore, no more than to disclaim belief.
14. Turner, having with Boykin made the first search in the river for Sheppard’s body, assisted in two subse*783quent searches, all of them fruitless- Iu the first and second long poles were used, and in the third, lead-lines and hooks. With this experience to qualify him, he was competent to testify, as he did, that “ it would have been a very difficult matter to recover a body in such a place.” The objection that this Avas giving an opinion with no reasons for it is not well-taken. The witness, in his preceding answers, had said, “Just at the bank where the boat lay there was a little counter-current about twelve or thirteen feet, but outside there was a very strong current; the water was very swift with the exception of a small counter-current near the bank. . . The water was about seven feet deep to a point about twelve feet from the bank, and from there out into the river it was about twelve or fifteen feet deep. There were some tree-tops and logs not over forty yards below the point where I found the boat. . . . There was current strong enough to wash the body under the logs and tree-tops just below, and, once under them, it is not probable that it would have got to the surface ” This detail of facts ought to satisfy even a captious demand for reasons.
15. In answer to one of the interrogatories addressed to him, Brown testified that -“ It was a strong current; the place showed to me as the place Sheppard fell in was open; just below, there were trees and drift-wood.” Besides an objection to the form of the interrogatory, the answer was objected to because it did not appear that the place shown to Brown was the correct place. The witness in answer to another interrogatory testified that a certain body of Avater (Moccasin Slough) runs into the river “ about three fourths of a mile above where Sheppard was killed.” This was objected to for substantially the same reason together with others. The same objection was urged to an opinion given by the witness that without a boat it would be impossible *784for a person to get out of the swamp, at the point where Boykin said Sheppard fell into the river. It is obvious that none of this evidence has any relevancy if the point to which the witness refers is not shown to be the one at which the boat from which Sheppard disappeared was found by Boykin and Turner. "We have searched the record in vain for something to vouch the accuracy and veracity of Brown’s informant as to the place where the boat was found. ¥e have also had the aid of counsel, both by oral argument and an elaborate brief, and our attention has been directed to nothing which supplies the missing link. The suggestion of counsel is that absence of the link goes not to the admissibility but to the effect of the evidence. It is clear, however, that no effect whatever is due to the evidence (and this is the test of relevancy) unless it applies to the place at which Sheppard disappeared. Of his own knowledge, Brown does not profess to know where the boat was found, or where Sheppard fell into the river, or that he fell in at all. A certain locality was pointed out to him by Boykin as being the place where Sheppard fell in, but unfortunately Brown gives no mark or description by which the place shown to him can be identified as one and the same with that which it was represented to be by his informant. He seems to have in view a point about three fourths of a mile below the mouth of Moccasin Slough, whereas Turner, the only witness who knew the true point, locates it about 600 yards below the mouth of the slough. The brief of counsel collects from the evidence mai’ks of identity as to the neighborhood, sufficient to show that the two points, if not the same, must be in the vicinity of each other. But this serves no purpose, for the simple reason that Brown, the witness, predicates the statements objected to, not of the neighborhood, but of a particular point, to wit, the “ place *785shown” to him,“where Sheppard was killed,” the “point where Boykin said Sheppard fell in.” He says that without a boat it would, in his opinion, be impossible for a person to get out of the swamp at the point where Boykin said Sheppard fell into the river; but whether it would or would not be possible at other points in the neighborhood he expresses no opinion ; nor was his opinion touching any other point called for by the interrogatories. The evidence, taken-as a whole, indicates that he was qualified to give a broader opinion than he did give, but the opinion he actually gave is restricted to the one point, and is consequently irrelevant, there being no evidence even tending to show that the point where Boykin said Sheppard fell in was the same at which he did fall in if he fell in at all. The opinion, as well as the other evidence of Brown touching the particular locality designated by his informant, should have been excluded.
16. The objection of irrelevancy which was taken to the whole of Laphan’s evidence was not sustainable. He was well-acquainted with the river, by actual experience as mate and pilot in navigating it, and testified as to its channel, current, etc., in the vicinity of the scene of Sheppard’s disappearance as located by the evidence of Turner. The information which he gives touching Moccasin Slough and the river Styx throws some light on what might be the nature of the current 600 yards below the mouth of the slough, where, according to Turner, the empty boat of Sheppard was found. In' connection with his opinion that a body falling overboard would probably sink to the bottom and be caught by and entangled with some of the numerous logs and tree-tops in the bottom, the rapid flow of the current holding it firmly there, the witness gave, besides his long experience in the navigation, the two facts that the water is very swift, and the bed of *786the river at that point is lined with large trees of various descriptions which have fallen from the bank. This relieves his opinion from the only objection to it which is specified in the motion for a new trial, namely, that he gave no reason for it. He undertakes to say as a fact that the bed of the river is lined with trees, and states nothing to show that this is mere opinion. Why should counsel ór this court assume it to be mere opinion? Nothing in the nature of the subject-matter renders it unknowable as a fact. That twenty years before Sheppard’s disappearance a person fell overboard at or near the mouth of Moccasin Slough, was drowned, and his body never recovered, was not admissible as an independent fact, but as a fact connected with the above opinion just given by the witness, was not without some slight relevancy. It was of little consequence for any purpose.
17. The witness Hochstrasser testified in answer to interrogatories as follows: “I am familiar with the Chattahoochee and Apalachicola rivers. I run on a steamboat on the two rivers for about five years.. The Apalachicola river was permanently obstructed by the Confederate forces and the water flowed through Moccasin Slough — a narrow passage from there into the river ‘Styx,’ returning to the Narrows of the Apalachicola river. The river Styx and the Chattahoochee run parallel with each other and Moccasin Slough connects them, and the Styx empties into Chattahoochee river about the Narrows. The swamp at this point is miles in extent, an average width of three miles, and impassable except in 'boats, and then by means of deep creeks and sloughs. The current of the river where Moccasin Slough runs into the river Styx just below extremely swift. That portion of the river is filled with rafts, logs, tree-tops and other obstructions. There is a very poor prospect of recovering a body that falls *787into the river just above these rafts or obstructions in a swift current, also very little prospect of recovering a body washed into the swamp by these cross-currents. There would be no chance to recover a body that fell into the river at the place mentioned, unless it should drift with the current of the river.” So much of this as relates to the prospect or chance of recovering a body in the river or the swamp was objected to as irrelevant, and as mere opinion without reasons, and because not referring to the place where Sheppard disappeared. In the light of Turner’s evidence it applies to the immediate vicinity of the disappearance, if not to the precise locality. It deals with the characteristics of river and swamp in that neighborhood. In the answer as sent here in the transcript of the record, there is evidently an omission of one or more words where the-witness says, “The current of the river where Moccasin Slough runs into the river Styx just below extremely swift.” Looking to the interrogatory to which this is a response, we find it as follows : “Is not the current of the river where Moccasin Slough runs into the river Styx, and just below, exceedingly swift, and is not that portion of the river filled with rafts, logs, tree-tops and other obstructions ?” Obviously, the answer was intended to be as broad as the question. So construing it, what is called this or that “point” is not restricted to a particular spot, as was kindred phraseology in the evidence of Brown {supra), but embraces the junction with Moccasin Slough and some indefinite portion of the river “just below.” In speaking of so large a river,, whether “just below” -would comprehend the distance of 600 yards, the distance indicated by Turner’s evidence from the mouth of Moccasin Slough to the spot where Sheppard disappeared, is not certain. The expression is ambiguous and could go to the jury for their construction. The *788witness gives facts on which, added to his general knowledge of the river, he might well base the opinion he expresses. Some assistance, though it might be very slight, could be derived by the jury from the facts and the opinion, taking them together. It seems to us that the evidence was not altogether without relevancy.
18. Sheppard resided in Liberty county, Florida, and was between 30 and 40 miles from home when he disappeared. After his disappearance, and in the latter part of the same month (January, 1885), a stranger calling himself Horton, which is the family name of Mrs. Sheppard, arrived at Perdue Hill, Munroe county, Alabama. He remained there most of the time until the following May; then left, as it would seem, via Selma and Montgomery, and the evidence affords no distinct trace of him afterwards. The defendant’s theory is that Horton was Sheppard. As tending to support this theory, it adduced the evidence of witnesses that early in May, 1885, they saw Sheppard upon the Alabama river, a passenger on the same steamboat with themselves, and that he landed from the boat at Claiborne, Ala., which is within 3 miles of Perdue Hill. The company also proved by other witnesses that an authentic photograph of Sheppard, exhibited to them at the trial, was, or seemed to be, a likeness of Horton. The plaintiff replied to this evidence by that of a witness, who, according to the brief of evidence, testified in substance that he knew Sheppard, had seen him twice, once in a business transaction about insurance, and was at his wedding; that in June or July, 1885, after he was reported dead, a man came to the hotel in Gainesville, Ga., and registered from Texas; witness watched him and thought he was Sheppard, but decided not to do anything until after making some investigation; found he had been living in Gainesville before and had just been off; got information to satisfy him*789self that this man was not Sheppard, and does not think his name was Horton. The motion for a new trial quotes (or rather misquotes) some of this evidence, and excepts to it as incompetent, because irrelevant and hearsay. The quotation runs thus: “In the summer of 1885, at Gainesville, Ga., he saw a man that looked so much like Thos. J. Sheppard, whom he had seen twice before, that he was satisfied that the man was Sheppard, and was about to have him arrested when he found out from the citizens of Gainesville that the man was an old citizen of that town.” Taking either version as correct, the sum of the evidence is that the witness saw a man whom he thought was Sheppard, but changed his opinion in consequence of information received from others. His informants, whether citizens or not, are anonymous. He does not testify that the man was or was not Sheppard, but that he first concluded that he was, and after investigation concluded that he wasn’t. He thinks his name was not Horton, but what he thinks it was, or what it was in fact, or what the citizens said it was, he nowhere discloses. If he had changed his mind on close and more careful inspection of the man whom he took at first to be Sheppard, the premises on which he proceeded as well as the final conclusion deduced therefrom would have been covered by his oath as a witness; but as the change was brought about, not by inspection, but by information derived from others, the premises are unverified save as to their effect upon his mind. All we know of them is that they were sufficient to convert him from one opinion to another; b\it for their actual truth there is absolutely no guaranty. It surely can be no answer to witnesses who think they are not mistaken upon a question of identity, to show that another witness was mistaken upon a like question in another instance, especially if the evidence of his mistake is chiefly, or even partly, what he has been *790told by third persons not sworn. Test the matter by a supposed case of handwriting. Witnesses swear that a lost deed is in the handwriting of Sheppard. Others come forward and depose'that they thought a certain other deed was in his handwriting, but on inquiry they became satisfied that it was not, but was in the handwriting of another person, not even naming him. No one would suppose such evidence admissible. Or take the still more apt illustration of a lost deed where the identity of the deed itself, not of the writer, is in question. Certain witnesses testify they saw the deed after it was lost, and others that they saw an original deed corresponding with a photographic copy produced in court. Would it be any answer to prove by another witness that he saw a deed which he at first thought was the lost document, but by reason of information which he considered reliable he became convinced that his opinion was erroneous, and therefore abandoned the measures which he had intended to take for the purpose of securing the forthcoming of the missing paper?
19. On the theory of answering the defendant’s proof that the photograph of Sheppard was a likeness of Horton, the plaintiff introduced three witnesses, each of whom testified that, on being shown the same photograph, he said to the person who presented it that it was the picture of Tom Collins, and three more who each testified that on a like occasion they said it was a picture of E. B. Smith. Even if what these witnesses thought at the trial as to the original of the photograph would be relevant, certainly what they said on a previous occasion as to their opinion then, had no relevancy The picture was an authentic photograph of Sheppard, and as evidence tending to show that Horton was Sheppard, the resemblance of the photograph to Horton was established. Nothing is more obvious than that this evidence could not be met by proof that certain persons *791said the photograph was like Collins, and certain others that it was like Smith. Moreover, grant that what they said was true, unless Collins or Smith was Horton, it is difficult to see how their resemblance to him would tend to prove that he was not Sheppard. Such an argument, if pursued to its consequences, could be used to prove that Sheppard was not himself. It is not within the range of possibility to negative, the accuracy of a witness who refers a copy to one original, by showing that divers witnesses were mistaken in referring it to another. His correctness would presuppose mistake on their part if they differed with him ; and if they were mistaken, he would have to differ with them to be correct. The logical result must be that their evidence would prove nothing, and would consequently be irrelevant.
20. Besides other policies of insurance upon his life, Sheppard had one payable to his brother. This was in the New England Mutual company. The brother testified at the trial that Sheppard intended to change that policy for the beuefit of his child, which was expected. He also testified that he (the witness) executed an assignment to the child because Sheppard declared such was his intention. The child.was bom in March, after Sheppard’s disappearance, and the assignment to it was executed in the winter following. The brother being a witness for the plaintiff, it was competent to show by him that he was not to be a gainer by Sheppard’s death, and that he was free from the bias of’ interest in giving his testimony.. For this purpose the fact of the assignment was relevant, and in explaining the transaction he could well state what induced it. The objection to so much of his evidence as consisted in reciting what Sheppard had said to him as to his intention, is not sustainable. The motion for a new trial complains that Mrs. Sheppard was allowed to testify to a statement made to her *792by her husband touching the same policy, and give her reply to the same, but the brief of evidence does not bear this out. The conversation detailed in the brief was not in relation to this policy but to another. It seems to us quite irrelevant, but as the motion for a new trial does not cover it, we forbear to deal with it as a point in the case. At most it is a mere trifle.
21. The plaintifl, pending the suit, it would seem, received a letter from an agent of the insurance company, which she did not answer because she considered it insulting. In her testimony she added to this explanation of her silence, that she sent the letter to her counsel, who advised her not to answer it. That this advice was given was objected to as hearsay and irrelevant. We consider it as a part of the explanation of her failure to reply to the letter, and therefore- admissible, if her explanation was relevant, and this seems to have been conceded, for some of it was not objected to. Even if error, we wonder that such a trifle should be supposed by learned counsel to be cause for a new trial. The same observation applies to many of the points made in this motion, they are so painfully minute and debilitated.
22. The code, §2850, subjecting insurance companies to liability for damages and reasonable counsel fees for bad faith in failing to pay losses, applies, both in its letter and purpose, to foreign companies doing business in this State, as well as to domestic companies. If it be true, therefore, that the policy on which this case is founded is a Connecticut, not a Georgia contract, and if it be true, also, that the laws of Connecticut do not impose such a liability, this policy having been issued by a company doing business in this State, is, nevertheless, subject to the provisions of our statute as to the consequences of bad faith in failing to comply with the company’s undertaking to pay within sixty days after *793demand. It follows that evidence of what would be a reasonable fee for plaintiff’s counsel was admissible. But that evidence should have been confined to a fee certain, inasmuch as no contract for a conditional fee was shown. The jury could have no use for evidence as to what would be a reasonable amount for any sort of fee which they could not allow, and surely they could make no allowance for a conditional fee, unless it was proved that there was a contract for such a fee. The law implies a contract between client and counsel for a certain, but none for a conditional fee. Nor was it competent to admit evidence as to any fee or compensation for services which would be rendered only on the contingency that the litigation should be protracted beyond the trial and verdict by motion for a new trial, or by writ of error. The jury could make no practical application of such evidence without assuming that the litigation would be so protracted, and that assumption might not be realized, in which event the company would have to pay for services not rendered or expenses not incurred. Any estimate of the witnesses, beyond the amount of a certain fee for services up to judgment on the verdict, and for the usual care and attention afterwards in enforcing the judgment, should have been excluded. But in order to carry out the statute in its full spirit and. purpose, it might, perhaps, be allowable in a proper case, for the jury, under the direction of the court, so.to mould their verdict as to leave the amount of fees for successful resistance to future litigation open for assessment by another jury after the litigation was all over. Under the code, §§3562, 4213, the power to mould verdicts seems broad enough to cover such a measure for the attainment of complete justice. The motion made to rule out all the evidence of the witnesses as to fees, because their basis of estimate included a misconception of the nature of the case, as involving *794the character of the plaintiff and the honor of her husband, was properly denied. The remedy was to have the witnesses, on cross-examination, to vary and correct their estimate by leaving out this element and reducing the amount accordingly.
23. We have now dealt with all the objections going to the substance of the evidence admitted, but it remains to dispose of objections which were taken to the form of the questions by which some of it was elicited. Two of the interrogatories propounded to Brown were attacked as leading, but as this exception was not filed with the interrogatories before the issuing of the commission, as required by the 37th rule of the superior courts (Code, p. 1350), it came too late. To thirty of the interrogatories propounded to Turner, an exception on this ground was duly filed, and was overruled by the court as to each and all of them. No doubt most of these interrogatories are leading, but many of them relate only to collateral or introductory matters not in dispute and not bearing directly on the merits of the controversy. As to such matters leading questions are allowable. 2 Taylor’s Ev. §1404; 1 Grlf’s Ev. §434.
But several of the interrogatories assume that Sheppard fell into the river, which was a disputed and most material fact, and one or two of them that he was dead, and that there was news of his death. This sort of assumption is one of the most- pernicious forms in which the vice of leading questions can make its appearance, its tendency being to induce the witness to adopt the theory of the facts propounded by the examiner, and shape his testimony in a way to lend support to that theory. Even an honest and well-meaning witness may sometimes be drawn by this device into coloring the letter, if not the spirit, of his evidence more highly than the exact truth, so far as his knowledge of it extends, would warrant. It is not lawful, as a general *795rule, to propound in chief “questions which involve or assume the answer which the party desires the witness to make, or which suggest disputed facts as to which the witness is to testify.” . 1 Wharf. Ev. §499; Steph. Dig. Ev. art. 128. That this rule was flagrantly violated in several instances will be manifest from a mere glance at the following interrogatories. The answers are also set out, not-to elucidate the question whether the interrogatories are leading, for on that question any answers would be irrelevant, hut to show that the letter of some of the answers was influenced by the form of the questions, although the witness did not know either that Sheppard was dead or that he fell into the rivei’, and made it perfectly clear elsewhere in his testimony that he had no such knowledge:
Q. “ State what was the nature of the current at the point where Sheppard fell in, whether it was slow or swift?” A. “Just at the point where the boat lay, there was a little counter-current, but 12 or 15 feet, about, outside there w.as a very strong current.”
Q. “ Ilow far below the mouth of Moccasin Slough was the point where Sheppard fell in the water?” A. “The point where Sheppard fell in the water was about 600 yards below the mouth of Moccasin Slough.”
Q, “State whether or not there were any logs or treetops or rafts, either where Sheppard fell in or just below, and if just below, how far below.” A. “There was some tree-tops and logs not over 40 yards where found the boat.”
Q. “ What time in the evening was it that Sheppard fell into the river, etc?” A. “It was very late in the evening, and very dark and cloudy. It was a dark and gloomy day. It was only a short time before night.”
Q. “Did either you or Boykin or Brown carry the news of Sheppard’s death that night to his wife or father and mother ? If you say you did not, why did *796not one of you three go ?” A. “We did not carry the news to the Sheppard family that night. I did not feel able to go.”
A few others of the leading questions were calculated to do harm, but perhaps did none, construing all the answers of the witness together. His testimony, in its general effect, makes the impression that he was a truthful and unbiased witness. In several instances the attempt to lead him was unsuccessful, and though the letter of his answers was sometimes shaped by the form of the question, there was no perversion of the spirit and meaning of his testimony, considered in its totality. For this reason we hold that the court did not abuse its discretion in admitting all the answers that were otherwise competent, notwithstanding the vicious character of many of the interrogatories. To admit or reject evidence drawn out by leading questions is generally discretionary with the trial court. Ewing v. Moses, 51 Ga. 410; Farkas v. Stewart, 73 Ga. 90; Parker v. Railroad Co., 83 Ga. 539.
(b) rulings rejecting evidence.
24. The company offered to prove that within two or three days after the disappearance of Sheppard, and before any insurance company knew of the same, it was currently reported and believed in the neighborhood of his home that he was not dead, but had run off to defraud the insurance companies; also, that this report was communicated, about the last of January, by the plaintiff herself and others, to the first insurance agent who visited the neighborhood. This evidence had three several purposes: (1) to repel the imputation of plaintiff’s counsel, announced to the jury, that defendant had put out this report, or one like it, in pursuance of a conspiracy with a certain witness, to fabricate a defence; (2) to vindicate the good faith of the company in refusing to pay; and (3) to corroborate *797the witnesses who had testified that they saw Sheppard alive after his disappearance. As to the first object, though the plaintiff’s counsel proclaimed a theory of conspiracy and fabrication and afterwards argued it to the jury, no evidence was adduced in support of it. This being so, what occasion was there to meet it otherwise than by a flat denial? It would expand and protract trials indefinitely, if evidence had to be heard in answer to all the ingenious theories advanced by counsel, or to forestall the invention and application of all theories that might seem plausible. To render evidence admissible requires relevancy to the case, to the matters in issue; relevancy to the sophistry, the invention or the elocution of counsel will not suffice. The second purpose, that of vindicating the good faith of the company, was no less wanting in sufficiency. We have already said, under another head of this opinion, that the question of good or bad faith must be solved by evidence applicable to the merits of the controversy, and not by a collateral inquiry depending on evidence having no relevancy to the merits but only to the special question. If the company thought proper to act upon a local report or rumor to the effect that Sheppard was still alive, it was responsible for tracing the rumor to its source and verifying its authenticity. An insurance company cannot justify its refusal to pay a loss in due time by pleading and proving a report or rumor that no loss had been sustained. The third purpose, that of corroborating witnesses who testified they had seen Sheppard, would seem to presuppose that cumulative evidence can dispense with some property or characteristic which it would have to possess were it offered without the evidence which it is designed to strengthen and fortify. Whilst this presupposition may be correct in some instances, we think this is not one of those instances. Unless the prevalence of the *798alleged report would be relevant evidence if offered by itself, as tending to prove Sheppard not dead but alive, we see not how it could gain admission by connecting it with what certain witnesses had seen. Their seeing was the same whether there was such a report or not; and unless the report has, in and of itself, some evidentiary force, the probability that Sheppard was in fact seen is neither more nor less than if no such report ever existed. The result is that the evidence offered and rejected was not admissible for any of the three purposes specified.
25. According to the motion for a new trial, the court rejected the evidence of the witness Revill in these words : “When Redden Boykin first saw me after the disappearance of Sheppard and told me about it, he did not appear like he felt all right, from the fact that he kept his eyes on the ground.” According to the brief of evidence (page 76 of the transcript) a part of the evidence of this witness received was in these words: “During the conversation, he (Boykin) did not appear like he felt all right, because he kept his eyes on the ground; he didn’t look towards me.” As Boykin’s interview with Revill was no part of the res gestee, what he said or how he looked or felt on that occasion was altogether irrelevant. It follows that not less, but more, of Revill’s testimony was received than was admissible. Furthermore, no court could be expected to grant a new trial upon this trivial and attenuated point, even were it well-founded in fact and in law.
26. Neil and Miss Huntwere witnesses for defendant. W. B. Sheppard was a witness for the plaintiff. Miss Hunt had written a letter to Neil’s wife, saying that she had seen Sheppard in Alabama. Neil had read a portion of this letter to W. B. Sheppard, and thereupon the latter had written to Miss Hunt, who returned no answer. "Whilst under cross-examination, W. B. Shep*799pard said he could not remember what Miss Hunt had written to Mrs. Neil, except that she had seen Sheppard’s ghost. The court refused to allow defendant’s counsel to read to the witness in the hearing of the jury any portion of the letter or any extract from it, as a foundation for asking him if that was not what Neil said to him. It refused also to allow the witness to read any of the letter aloud before the jury, and would only allow him to read it silently ,for the purpose of refreshing his memory. It also rejected the letter itself as evidence. If the witness, in writing to Miss Hunt, acted upon her letter to Mrs. Neil, it was only indirectly, for he testified that he had not read any of the letter. He acted, therefore, not upon the letter itself, but upon what Neil read as a portion of its contents. He was allowed to repeat to the jury from memory all that Neil read, and to use the letter to refresh his memory not only as to the sense, but the words of the reading to which he had listened. Surely this was enough to explain the relation of the letter to the mere fact of waiting a letter on the same subject to Miss Hunt, and it was only this fact and the failure of Miss Hunt to reply which the court thought relevant, the bearing of the testimony being, as we may suppose, to illustrate the bias of Miss Hunt as a witness in the case for the defendant, at whose instance she had come from Alabama to attend the trial, and yet had failed to so much as answer a letter written to'her by the brother-in-law of the plaintiff The court held, not only that one of these letters, but that both of them when offered by the defendant together, were irrelevant, neither being offered for the purpose of impeaching the witness ~W. B. Sheppard. Certainly, letters from one witness to a third person, and from one mere witness to another, whatever might be their contents, ought not to affect either of the parties to the suit otherwise than as a *800means of disci’editing one or both of the correspondents. Of course, if the letters were inadmissible, any examination touching their contents which would involve a direct recital of such contents would be, not equally, but more inadmissible; consequently it was not error to deny defendant’s counsel the privilege of referring to the contents of the letters in cross-examining the witness. As the object, according to the record, was not impeachment, it matters not what other object was in view. There was no evidence on which to found any theory that the witness was in combination with his brother to defraud the insurance companies, though that theory was formulated and insisted upon.
27. Pending this suit, a letter was written to the plaintiff by an agent of the insurance company, requesting certain information for use in identifying her husband, and inviting her co-operation in further efforts to find him. This letter, to which no reply was made, was ruled not to be admissible evidence in behalf of the company. The correctness of this ruling is self-evident.
28. On cross-examination, a witness was asked this question by defendant’s counsel: “I understand you to say when Mr. Rutherford examined you, in your opinion what about the writings of those papers — those letters ? (referring to the letters signed ‘Murray’ and ‘T. M. Horton’)” It is not stated what objection, but only that objection was made to the question and sustained. This ruling is complained of as error. Perhaps the objection was that the question was nonsense, or at least unintelligible; and if so, it was well-taken. Manifestly, the question as it comes to us in the record, whatever may have been its previous state and condition, is amenable to that objection. It seems to have a “glimmering of reason,” but no more.
(c) NATURE OE THE PLAINTIEE’S EVIDENCE, AND ITS SUEEICIENCY TO WITHSTAND A MOTION EOR NONSUIT.
29. A motion for a nonsuit was made upon two *801grounds: (1) that the application for insurance was not put in evidence with the policy; (2) because the testimony did not establish, in accordance with the terms of the policy, by direct and positive proof, that Sheppard was dead, or if dead, that it was from a personal injury caused by external violence and accidental means. The motion was overruled. Touching the first ground, we have already dealt with it substantially under the first head of this opinion, the conclusion there announced being that it was not incumbent on the plaintiff to introduce the application, and that the policy might he regarded, for the purposes of the plaintiff’s case, as setting forth the contract on which the action is founded. The principle involved in the second ground of the motion has also been dealt with, incidentally and to a limited extent, under the third head of the opinion where the terms of the policy with reference to preliminary proof were considered and discussed. The provisions of the policy directly involved by the present question read as follows:
“Provided always, that this insurance shall not extend to . . any bodily injury of which there shall be no external and visible sign, . . nor to any case except where the injury is the proximate and sole cause of the . . death. And no claim shall be made under this policy when the death . . may have been caused by suicide, felonious or otherwise, sane or insane. And this insurance shall not be held to extend to disappearances, or to any case of death or personal injui’y, unless the claimant' under this policy shall establish by direct and positive proof, that the said death or personal injury was caused by external violence and accidental means.”
The company sets up no affirmative defence involving death by suicide or any other means not covered by the policy. The fact of death by any means is controverted, and the continuance of life is affirmatively predicated. Still, the plaintiff must make out her case; but in so *802doing she can use the presumption against suicide which the law recognizes as arising out of the instincts of nature, one of which is the love of life. "Where the fact of death is established, and the evidence points equally or indifferently to accident or suicide as the cause of it, the theory of accident rather than of suicide is to be adopted. Cronkhite v. Trav. Ins. Co. (Wis. 1889), 48 N. W. Rep. 731, 19 Ins. L. Jour. 267; Mallory v. The Same, 47 N. Y. 52; Trav. Ins. Co. v. McConkey, 127 U. S. 661. To these three cases the same insurance company which is now before us was a party, and the provisions of the policies involved in them were no doubt the same as in the one upon which we are adjudicating. Any construction of these terms which would dispense with direct and positive evidence as between suicide and accident, would, if carried out consistently, dispense with it as to the fact of death itself. And we think such a construction, if possible, ought to be and must be adopted and applied. In one sense all oral evidence is direct. Indeed, a very eminent authority on the subject takes no account whatever of circumstantial evidence, but treats the one sense of “ direct ” to which we have referred as the only sense in which the word is correctly used in the law of oral evidence. He says, “Oral evidence must in all cases be direct.” Stephen Dig. Ev. art. 62. His treatment of the subject shows that what he means is that, whether the fact to which a witness testifies be the ultimate fact to be proved or only an evidentiary fact tending to prove it, the testimony itself must apply immediately and directly to the fact as to which the witness speaks. If we take the policy as meaning, by the use of the terms “ direct and positive,” this much and no more, it can be easily reconciled with the rules of law on the subject of evidence. But the more general, and with us the established use of the terms “ direct evidence ” gives to them a different *803and more restricted meaning. “All evidence may be divided into two classes: (1) direct, which consists in the testimony, whether immediately or mediately derived from those who had actual knowledge of the principal or disputed fact; or (2) indirect or inferential evidence, where an inference is made as to the truth of the disputed fact, not by means of the actual knowledge which any witness had of the fact, but from collateral facts ascertained by "competent means.” 1 Stark. Ev. 15. Our code, §3748, says : “ Direct evidence is that which immediately points to the question at issue. Indirect or circumstantial is that which only tends to establish the issue by proof of various facts sustaining, by their consistency, the hypothesis claimed.” And §3750 declares : “ Generally the rules of evidence are the same in all the courts of this State and upon every trial; the exceptions exist only by express statute.” It is manifest that, if it was the purpose of the insurance company to exact by contract direct evidence of death and accident, in the sense contemplated by Starkie and our code, as above quoted, then the rule of evidence laid down by the policy is in conflict with that established by law, a conflict which must be reconciled by adherence to the law and repudiation of this article in the policy. The rule which we must observe, where no statutory exception can be shown, is that all questions, civil and criminal, can be solved by circumstantial as well as by direct evidence, and that in civil eases the preponderance of testimony is considered sufficient to produce mental conviction. Code, §3749. Whether that preponderance is the result of one or the other kind of evidence or of a union of both, our law treats it as decisive of the case. And no matter where a contract is made or a cause of action arises, the rules of evidence applicable to the litigation are controlled by the law of the forum. The general principle that legal *804rules of evidence are not controllable by mere stipulation or contract, is a settled question with us. Carrugi v. Ins. Co., 40 Ga. 141; Ins. Co. v. Carrugi, 41 Ga. 674; Ins. Co. v. Coleman, 58 Ga. 255. And we understand the principle to be recognized elsewhere. “The condition that direct and positive proof must he made of death having been caused by external, violent and accidental means, did not deprive the plaintiff, when making such proof, of the benefit of the rules of law established for the guidance of courts and juries in the investigation and determination of facts.” Trav. Ins. Co. v. McConkey, 127 U. S. 661. “The evidence was sufficient within the rules of law; and the language of the policy cannot be construed to take the case out of the ordinary rules of evidence.” Reynolds v. Ins. Co., 1 N. Y. Supp. 738 (June, 1888).
Though the whole of plaintiff’s evidence is circumstantial, and though the facts directly in issue are established, if at all, inferentially from circumstances, the evidence was sufficient to entitle her to go to the jury, and the motion for a nonsuit was properly denied.
(d) CONDUCT OE THE TRIAL, AND IRREGULARITIES PENDING ITS PROGRESS.
30. One of the questions which arose during the trial and which bore on the alleged identity of Horton with Sheppard, was whether certain letters, purporting to be written by Horton, were in the same handwriting as certain other documents, which latter were written by Sheppard. On this question experts were examined by defendant, and to test their skill the plaintiff’s counsel interrogated them upon certain incomplete or mutilated writings which were partial copies by another hand of documents 'which Sheppard had written, though that the copies were written by another and not by Sheppard was not avowed or proved until this test examination was concluded. These copies being some of the *805writings to which the expert testimony applied, and being proved to be copies not written by Sheppard, were admitted in evidence, to which defendant objected because they were not proved to be genuine, nor submitted to defendant before the announcement of ready, and because they had been mutilated by plaintiff’s counsel in presence of the court by cutting off matter which was neither submitted to defendant’s counsel nor tendered in evidence. The purpose of the experiment was to show that the experts were not reliable; that they would mistake writing not done by Sheppard for writing which he had done. The whole object of the test would have been defeated, had other proof been required that the documents were genuine before they could be used. By what color of right could the defendant claim to see or know anything of them before the trial commenced? And why should not scraps and fragments be used to test the experts on the mere question of manual or mechanical execution? The contents or completeness of the writings had no evidentiary relation to the hand in which they were written. The complaint that plaintiff’s counsel was allowed to fold the papers and exhibit to one of the experts only the part exposed, and the further complaint that he was allowed to exhibit them to him at all for comparison with the genuine writings or other writings attributed to Sheppard, are both without merit, for the reason that to test the skill of the expert, it was reasonable and proper to grant to counsel a wide and liberal discretion in the use of means and methods. The same observation applies to cutting off and withholding such parts of the papers as counsel wished to conceal, the papers being his own and not then having been offered in evidence.
31. Defendant’s counsel, in the re-direct examination of one of his own witnesses, commenced a recital as fol*806lows: “I understand you at first to decline to give an opinion upon the letters of June 2d and February 5th, because one was written in pencil; but I understood you afterwards, when Mr. Rutherford examined you, to say — ” Here the court, on objection of plaintiff’s counsel, refused to allow the defendant’s counsel to continue the recital of his understanding of what the witness had said on the cross-examination, and to call for an explanation thereof. It was only the recital which the court arrested, not any explanation offered by the witness nor any direct question adapted to elicit an explanation. Certainly it was not error for the court to cut counsel short in a recapitulation moulding and shaping the testimony in his own language. If a question to draw out an explanation was intended, it could have been' put without the elaborate introduction which the counsel’s opening foreboded. ¥e cannot think it a proper practice for counsel to deliver an exordium on his understanding of what a witness has said, to clear the way for an explanation. Counsel knows what his witness has said, or does not know, or is in doubt. If he knows, he can put his question at once; if he does not know, or is in doubt, he can ask the witness to tell, and then put his question. In this way an explanation can be obtained, with no recital whatever of counsel’s understanding, and it is the better, and usually the briefer way. At all events, the discretion of the court as to such matters is ample, and in this instance, we doubt not, it was properly exercised.
32. There was a letter from one of defendant’s female witnesses to the plaintiff, which the defendant wished to use in evidence. Notice to produce it had been given. By consent of parties plaintiff’s counsel retained possession of it, but with a pi’omise to produce it when desired. By like consent its production was further postponed; but at length, whilst the writer was *807on the stand as a witness, defendant’s counsel,- desiring to exhibit the letter to her for identification, and then to examine her upon it, and finally to offer it in evidence, called for the letter, and plaintiff’s counsel refused to surrender it. "Whereupon the court was moved to compel its pi’oduction. Plaintiff's counsel responded that, because the matter was delicate, he preferred not to give his reasons in the presence of the writer of the letter why it ought not to change custody, and suggested that the writer retire. The court inquired if there was any objection, and defendant’s counsel replied that if any attack was to be made on her, it was her due to remain and hear it, and further stated that she intended returning to Alabama that night. The court then ruled that the motion to compel production was not in order, and postponed the matter until next morning. The witness was in attendance upon the court during the following morning, and did not leave the city until two o’clock in the afternoon. There is no complaint that the letter was not produced in time to serve every purpose for which it was desired, but the defendant wants a new trial because the court erred in ruling the motion out of order, and in postponing the matter until next morning. Doubtless, according to strict pjractiee, the motion was out of order; but if the court erred in so ruling and in declining to entertain it until next morning, what harm was done for which a new trial should be granted? An immaterial error is the same as none.
33. In opening the case to the jury, counsel for plaintiff had charged a conspiracy hy the defendant company and two of its witnesses to make it appear falsely that Sheppard was alive. One of these witnesses resided in Alabama, and had taught school in Florida. "Witnesses brought by defendant from these States were in attendance upon the court to support her character. *808When her examination was concluded, the trial had been in progress for ten days, and the non-resident witnesses had become impatient to return to their homes, and had announced their determination to leave on the next train, if they could not be examined that day. Because it was likely that several days more would be consumed in taking testimony before these witnesses could be called in regular order, counsel asked leave to examine them then, that is, immediately after the witness concerning whose character they were to testify; which leave the court denied on the ground that no impeaching evidence had been adduced by the plaintiff, and therefore no supporting evidence was at that stage admissible. The result was, that the supporting witnesses returned to their homes before the plaintiff' introduced any further evidence, and were not examined. The reason given by the court for this ruling is one of the commonplaces of the law. The rules of law cannot be varied to serve the convenience or even the necessities of witnesses, resident or non-resident.
34. Several letters, as one after another they were proved to be genuine, were handed by defendant’s counsel to plaintiff’s counsel for inspection. On handing them back three of them were pointed out by the lattei', with the request not to mix these three with the othex-s; to which defendant’s counsel replied that he would attend to them. Thereupon, at the instance of plaintiff’s counsel, the court ruled that the counsel had a right to have the papers identified so that they could not be mixed with other paper’s, and said that counsel for defendant could put some mark upon them for identification. The counsel refusing to do this, the coux't, over counsel’s protest, directed him to hand them to the stenographer to be xnai’ked, and directed the stenographer to mark them. When this occurred the letters had xxot been offex’ed in evidence, nor had the *809direct- examination of the witness called to prove them been concluded. In appealing to the court to intervene in the matter, counsel for plaintiff said he wished the letters put in the hands of the stenographer or the court so that he could cross-examine the witness upon them, and that “they” (meaning apparently opposing counsel and client) might get them mixed up with the others so that he could not have the privilege of examining the witness concerning them. In our opinion, the action of the court was not only an unwarrantable interference with the custody of private papers, they not having then been offered in evidence so as to become “assets of the case,” but that, though not so intended, it was a mischievous interference, iu that it would naturally be construed by the jury as implying a suspicion in the mind of the court that the defendant’s counsel could not be trusted to keep his own papers as they ought to be kept and deal with them fairly without being watched. The counsel could mark the papers himself, or let the stenographer mark them, but they must be marked. For what? For identification lest they should get mixed with other papers, and the right of cross-examination touching them be endangered or lost. Suppose they should have gotten mixed, and become incapable of identification before cross-examination of the witness, no harm could have resulted to the plaintiff. The letters could not have been put in evidence on the unfinished examination of the witness, and the risk of preserving their identity and the means of proving it was upon the defendant. As they had not needed any special mark for identification before they made their appearance at the trial, why should they need any afterwards until they had become “assets of the case”? A court should not on mere motion aid one side to police the other as to papers of which the court has no custody, but which *810are in the exclusive custody of one of the parties or his counsel. Such interference during the trial is calculated to impress the jury unfavorably towards the side over which the surveillance is exercised.
35. Certain letters were offered in evidence by the defendant, to which plaintiff’s counsel did not object, and so announced; but he proceeded to make an explanatory speech on the subject, to this effect: “ To the two letters of June 2d we could object, and are unwilling for them to come in without a statement; for that might put us in the position of admitting that they have been proved in the proper way. I will state my objection. These two papers purport on the inside to be addressed to ‘ My dear Emma,’ but there is no proof that they were received by the person to whom they were addressed — no proof that these identical papers were received by her. Mr. Roberts refused to say how he got them. When my brother Davis objected I was in the act of stating that we did not want it taken for granted, from our allowing these papers to come in, that there had been any proof that these were the identical papers. I think I have the right to make the statement to the court that they have not been proved; and I believe we could successfully object to them, but as we want them in, I will allow them to come in. He (Mr. Davis) raised a question of the impropriety of the. objection, but there is no evidence that this person ever received these letters, and none that it is the identical handwriting which came to the person to whom they were addressed. That would be a substantial objection, and we could urge it and rule out any part of this testimony.” The court, over the objection of defendant’s counsel, allowed this speech to be injected into the evidence as it came in. So far as we know, such a practice is without precedent. It is in a high degree aggressive and improper. A verdict won *811by such means in a cáse fairly doubtful or disputable ought to be set aside. Misconduct of counsel tending to prejudice his adversary is inconsistent with fair trial, and it is a, miscarriage of forensic j ustice for a suitor to lose his case without a fair trial, whether he deserves to lose it when properly tried or not. Counsel are no less bound than courts not to violate the conditions of fair trial. If they do it, they and their clients must take the consequences, and must understand that the fruits of success in an unfair trial are not secure. In the present instance, we are altogether certain that counsel had no conscious intention to break through the proprieties, much less the principles, of trial; nevertheless, led no doubt by his zeal, he did both. He announced, in substance, that he did not object to the evidence offered. If he did not, what right had he to address the court in the hearing of the jury on the infirmities of the evidence, and on its liability to an objection which he did not propose to use to exclude it ? The reason he assigned was, that without making a statement to the court, he might be understood as admitting something which he denied. But why should he not, if the trial was to be a fair .one, be understood as admitting all that his failure to object to the evidence legally implied, and what danger was there that he would be understood as admitting more ? He was not entitled to be heard in such a speech at that stage of the trial to save his conduct in forbeai’ing to object to the evidence from any misconstruction whatever. He was out of order, whether the admission which he wished to bar flowed from his conduct or not. If it did, he had no right to bar it; and if it did not, there was no need to bar it. Thus the speech had no legitimate function. "What was its illegitimate function ? To disparage the evidence as it passed in on its way to the jury; to taunt it as illegal, because amenable to a substantial objection *812which the counsel might urge and have it excluded, but which, partly of his free grace and favor, and partly because he wanted the evidence in, he passed by without so much as submitting it for the opinion of the court. Had he taken that opinion, the court might have differed with him, and so it might be best to decide the question himself and proclaim his decision in an informal address to the court. Moreover, he himself wanted the evidence in, and if so, what sense or reason would there be in objecting to it ? Why did he want it in ? Of course, because it would not only be harmless but beneficial to his side. This not only runs it down to zero, as evidence for the defendant, but below zero. What more could be said to disparage evidence, than that it is too weak legally to get before the jury, and so unserviceable after it gets there that the other side wánts it in, and though able to keep it out forbears to do so?
36. At the request of plaintiff’s counsel, defendant’s witnesses were “put under the rule/’ that is, each witness, until his own examination, was required to remain out of the coui’t-room and out of hearing of the testimony of other witnesses. Whilst one of the witnesses was under examination, plaintiff’s counsel complained in open court and in the presence and hearing of the jury that Hr. Lewis, an agent of defendant, who was assisting in the trial of the cause, went out after the examination of some of the witnesses, and talked to others who had not yet been examined; that these conversations were private, and what was said the counsel did not know; that his information was derived from the sheriff in charge, whose duty it was to look after such matters, and that he, the counsel, wanted the practice stopped ; that it was wrong, and he could see no necessity for the agent going and talking with his witnesses immediately after others of them were examined. Counsel for defendant replied that he had Hr. Lewis’ *813authority for saying that he had mentioned to none of his -witnesses anything that another had testified. Plaintifi’s counsel replied, “It would be better for him to state it under oath.” The court had a public investigation in presence of the jury, and found that Dr. Lewis had gone out and conversed with another agent of the company, who was “under the rule,” but had no improper conversation. Of course, it is not allowable for counsel to raise a side issue, pending a trial, which casts a groundless imputation upon the good faith of the adverse party in dealing with witnesses over whom the court has assumed a special control for the very purpose of keeping their testimony pure. Any public charge or insinuation of tampering with sources of evidence tends to render the accused party and his case odious, and it is reckless mischief to- put such poison afloat in the atmosphere of a trial without good cause for it. Even with cause apparently good, a considerate regard for the possibilities of mistake would suggest that such matters should not be opened and discussed in the hearing of the jury. Opportunity can always be made or found to obtain the ear of the court for so delicate a topic without having the jury for an audience. The jury, if necessary, can be directed by the court to withdraw, and a request for such withdrawal ought to be made before broaching the subject. The jury have .no concern with the conduct of the parties not falling-under their own observation, unless the conduct is of such a nature as to be evidence for their considei-ation in the pending trial; and if of that nature, it should be proved before them as part of the evidence in the cause. No fact which is to influence their verdict should reach them otherwise. By any test of wholesome practice, though the letter of counsel’s complaint as to the mere fact of intercourse with one of the sequestered witnesses was well-founded, the complaint *814had too much of the “hue and cry” in it. Looking to the time and manner of making it, it seems to have had, relatively to the jury, more publicity than was needful. This is all the criticism to which we think it exposed as to method. • As to matter, confining the question to the letter of the complaint, there may be doubt whether the facts directly charged amounted to any violation of the judicial quai'antine. When witnesses are sequestered, intercourse with them for a purpose inconsistent with the object to be subserved, is, of course, interdicted; but whether total non-intercourse without leave of the court is enjoined, unless there is some express order to that effect, is not quite certain. The distinction between the two constructions of the rule is material, for it would not be consistent to hold that intercourse is free for all purposes but one, and then hold that the bare fact of intercourse affords just ground for complaint so as to put the party on a vindication of its object by purgation or proof. The presumption would be in favor of the object being legitimate, not against it. As we have listened to no careful discussion of the question, we leave it undecided. This we can do with propriety, as in disposing of the present case we are uninfluenced by this ground of the motion for a new trial.
37. One of the grounds of the motion for a new trial imputes misconduct to a deputy-sheriff who was in attendance upon the court, and who, during the progress of the trial, made some verbal demonstration of feeling in favor of the plaintiff and against the defendant, and some of the defendant’s witnesses, who were waiting at the time to be examined, having been sequestered by order of the court, may have heard his discourse. This ground, as it stands in the motion, is not fully verified by the judge. We will only deal very briefly with the principle involved in it, and need not even recite *815the particular facts brought, out against the conduct of the officer. It is enough to say that the language imputed to him might have some tendency to depress and discourage, if not overawe witnesses of susceptible feeling and timid disposition. Of course it is highly reprehensible for an officer to manifest partisanship in a case on trial in a way to produce that kind of result. Instead of being a minister aiding in dispensing justice, he would be an agent of wrong and injustice. In the present case, however, it does not appear that the officer wrought any actual mischief, though he may have attempted it. Ilis misconduct was not complained of, and the court made no ruling upon it, until after verdict. This being so, in order for it to be cause for a new trial, some actual injury done by it would have to be shown.
38. Having discussed and disposed of all the special grounds of the motion for a new trial, it remains only to touch upon the general grounds. And in doing this we shall confine our observations to that element of the verdict which burdens the defendant with damages and counsel fees, leaving the recovery of principal and interest to abide the result of another trial, with no intimation from us as to the sufficiency or insufficiency of the evidence, taken as a whole, to warrant such recovery. We have a conviction, to the extent of absolute certainty, that the allowance of damages and counsel fees was groundless and unwarranted. Any rule or principle which would deny to the company the right of full and free litigation in such a case as this would deny it in any case. The company had stipulated in the policy against liability in the event of suicide or mere disappearance. Here was a disappearance, and whether, in the light of all the evidence, it was the result of death or of device, is fairly questionable. Two witnesses who knew Sheppard testify positively that *816they saw him after his disappearance. The man calling himself Horton turned up in Alabama within less than a month after the disappearance. The coincidences tending to identify him with Sheppard are numerous, such as (1) personal resemblance, including photographic likeness, complexion, color of hair and eyes; (2) resemblance of handwriting; (3) each had a process for clarifying syrup; (4) each had studied, or professed to have studied medicine; (5) the name Horton was Mrs. Sheppard’s family name before her marriage. On the other hand, the evidence tending to support the theory of death is very strong, and may be quite sufficient to resist and overbear all that which has a com trary tendency. But these conflicting tendencies ought to be determined by a jury, and it is no' abuse of the privilege of litigation for the insurance company to await such a determination and forbear to .make a decision against itself. As well might the plaintiff be required to make a decision against herself on pain of damages and counsel fees in case it should turn out that a jury should believe her husband is not dead and so find.
Though we have pointed out several errors committed by the court in the progress of the trial, most of them are of such minor consequence as to be no cause for granting a new tidal. But á few of them had a material bearing, some on the merits of the controversy, and one' or two on the right of the company to have a fair trial. The result is that a new trial is necessary.
Judgment reversed.