Ladd, J.
i. fruusbossd tiouslpiiysiuentfexpert testimony. All the errors assigned, save those with respect ■to rulings on the proposed testimony of Dr. Euml, and as to the sufficiency of the evidence to support the verdict, seem disposed of by previous decisions of this court, Li view of another trial, we refrain, as is our custom, from discussing the quantum of the evidence. But see Smith v. Sioux City, 119 Iowa, 50. The plaintiff testified that her right foot caught in the sidewalk and threw her on the side back toward the wall, She ivas then forty-seven years old. Her ankle bésame somewhat swollen and discolored. There was no wasting of the flesh of the arms or legs, but the evidence tended *49to show that the right half of the body became partially paralyzed. One of the physicians in attendence testified that in his opinion the paralysis was caused by the fall and in this view another, in answer to a hypothetical question, concurred. Four physicians were called in behalf of the defendant, three of whom expressed an opinion, in response to hypothetical questions reciting the facts as above stated, that the paralysis could not have resulted from the fall, but was due to some central or brain trouble or disease. When Dr. Euml was called as a witness, the parties agreed that he had been one of the attending physicians of plaintiff, and no exception is taken to the exclusion of any testimony by him touching his relations to her as a physician.. It was also agreed that the record should stand as if the questions propounded to the other experts had been asked him, and the objections separately interposed. The court then sustained the objections. Hypothetical questions, without referring to or disclosing the witness’ former employment as her physician, were also held incompetent, under section 4680 of the Code, providing that “no practicing * * * physician * * * shall be allowed in giving testimony to disclose any confidential communication properly entrusted to him in his professional capacity and necessary and proper to enable him to discharge the functions of his office according to the usual course of practice or discipline.” But no communication, confidential or otherwise, was sought to be elicited, and any intention to attempt'this was expressly disclaimed. The question did not refer thereto, directly or indirectly. Manifestly, then, the statute did not authorize the exclusion of the testimony. The record contains no suggestion of the physician’s inability to disássociate the facts stated in the questions from what he had learned from his patient. Indeed, the nature of the inquiry'seems to obviate any such difficulty. If, as counsel for appellee argue, the jury might infer from the fact he had treated plaintiff that “he knew more about the case than the experts, and -would be likely to give great weight to his evi*50deuce,” this would not justify the rejection of his testimony, but, rather, indicate the necessity for an instruction directing that no consideration be given to such fact. An unfavorable inference is always to be drawn from the omission of a party to call an available witness to the facts in dispute. The failure of an injured person to examine the attending physician as to the nature and extent of his injuries forms no exception to the rule. This, we think, marks the distinction between such physician and others, when testifying as exports in response to purely hypothetical questions. It is legitimate matter for the jury’s consideration, and not to be obviated by any rulings of the court. Our conclusion finds support in People v. Schulyer, 106 N. Y. 298 (12 N. E. Rep. 783; Valensin v. Valensin, 73 Cal. 106 (14 Pac. Rep. 397). See Herries v. City of Waterloo, 114 Iowa, 377; McConnell v. Osage, 80 Iowa, 293, is not in point:
z. cumulative evidence: exclusion. But appellee insists the ruling was without prejudice, for that three experts were examined, and the testimony, if received, would have been cumulative in character. The court had- not limited the number of experts which might be called on each side. Moreover, the mere fact that evidence is cumulative in character will not justify its exclusion. If evidence is cumulative to other undisputed testimony, its rejection has been held to have been without prejudice. State v. McPherson, 114 Iowa, 492; Morgan v. Wilfley, 71 Iowa, 212. But where there is a sharp conflict in the evidence, as in this case, the number or character of the witnesses often control, and the erroneous rejection of the testimony of any cannot be held, as a matter of law, to have been without prejudice. The motion to strike is overruled. — Reversed.