I dissent. If the policy here involved were drawn in terms of causation, I would have no doubt that the result reached by Judge BOWEN in his opinion is correct. The language of this policy, however, is peculiar. It insures against “bodily injury which shall occasion death,” and that phrase is expressly defined as follows:

It will be noted that drowning resulting from the causes defined is covered. The particular exclusion with which we are concerned is contained in a paragraph that lists a number of exclusions. The precise language is: “This Certificate does not cover death * * *

It will be noted that exclusions (a) through (d) are all cast in terms of causation. Exclusion (e), the one here involved, is not so cast. It applies “while.”

Bearing in mind the familiar rule that a policy is to be construed against the insurance company which prepared it, a rule which is reinforced in this case by a second rule, that exclusions are to be narrowly construed (see Brown v. Underwriters at Lloyd’s, 1958, 53 Wash.2d 142 , 332 P.2d 228 ), I am of the opinion that the case before us should have been given to the jury.

Both the trial judge and Judge BOWEN conclude that the evidence would per *533 mit a finding that the decedent did not suffer a fatal injury while he was operating the airplane; that is, up to the time when the plane had finally come to rest at the bottom of the lake. The physical evidence is, as Judge BOWEN states, undisputed, but more than one inference can be drawn from it. It is for the jury to draw inferences. Here it could infer that the decedent got out of the plane and clear of it, and that his death occurred thereafter, by drowning. I also think that if the jury should so find, the decedent would no longer at that time have been “operating or serving as a member of the crew” of the plane. There is a time at which such activities cease at the end of a flight. Because the policy was drawn by the insurance company and because the language is exclusionary, I think that we should hold that that time is when the decedent got out of the airplane, even though the peril to which he was subjected by reason of the plane’s landing in the water was still present.

I note that in the case of Wendorff v. Missouri State Life Ins. Co., 1927, 318 Mo. 363 , 1 S.W.2d 99 , 57 A.L.R. 615 , the policy language is quite different. There the exclusion not only covers injury sustained while in or on an airplane, but also injuries sustained in falling therefrom or therewith. No such language appears in the present policy. It would appear that exclusion (e) was deliberately drawn to eliminate causation as a factor determining whether the exclusion applies. It is easy to say that the drowning which the jury might find to be the cause of death was “directly or indirectly caused or contributed to” by operating or serving as a member of the ■crew of the aircraft and there would be no doubt about the proper result in this case if exclusion (e) had been so phrased. The problem is that it was not so phrased and presumably, particularly since it is a part of the same paragraph of the policy, the difference in phrasing was intentional.

I agree with the conclusion that the Coroner’s certificate was properly excluded from evidence. The record shows that it was filled out by a Dr. Butterworth, who had no knowledge of the cause of death, but obtained his information from a Mr. Ward, who was not qualified to express an opinion on the subject. Ward appeared at the trial and the court found that he was not so qualified. Thus, this particular certificate was the rankest kind of hearsay.

I would reverse.