Defendant Eli Lilly and Co. (Lilly) appeals from a judgment below awarding plaintiff Andrea Goldstein $50,000 for injuries she claims were caused by a drug, diethylstilbestrol (DES), ingested by her mother while Andrea was in utero. A separate claim of Mrs. Goldstein's husband, Paul, for loss of consortium was denied by the jury and no appeal has been taken from it.

In August 1977, Andrea and Paul Goldstein were added as plaintiffs to the case of Payton v. Abbott Laboratories, which had begun in April 1976. This case was begun as a class action and was conditionally certified, but subsequently the plaintiff class was decertified and the plaintiffs' suits were separated for individual trials.

The issues are: (1) whether the district court erred in striking one of the jury's special findings and not ordering a new trial; (2) whether there were inconsistencies in the jury's answers to interrogatories sufficient to require a new trial; (3) whether alleged errors in the jury charge require a new trial; (4) whether the alleged failures of proof on causation and/or proof that defendants manufactured the DES taken by plaintiff's mother require reversal and judgment for the defendant.

In the early 1950's, when plaintiff Andrea Goldstein was born, DES was sometimes prescribed according to the Smith & Smith regimen1 for pregnant women thought to be in danger of miscarriage. Andrea Goldstein's mother, Mrs. Schwartz, besides having one healthy baby boy also had two miscarriages before becoming pregnant with Andrea. Apparently because of this history of prior miscarriages, Dr. Ko, a resident at the New England Hospital where Mrs. Schwartz was given prenatal care, prescribed the Smith & Smith DES regimen for her beginning in the ninth week of pregnancy. The regimen provided for continually increasing amounts of DES to be taken by the expectant mother, beginning with a daily dose of 5 mg. in the seventh and eighth weeks of pregnancy, and increasing by 5 mg. every other week until the fifteenth week, after which it was increased by 5 mg. each week until the thirty-fifth and final week when the daily dosage was 125 mg. In order to make the daily ingestion of the increasing amounts of the drug easier for the expectant mother, it was recommended that 5 mg. tablets be taken in the seventh through the fourteenth weeks and 25 mg. tablets be taken thereafter. Although Mrs. Schwartz could not remember the brand of DES she took and the original prescription could not be found, she did remember she had had the prescription filled at Trachtenberg & Meyers Pharmacy, and that the first pills she took (the 5 mg. ones) were red and the second ones (the 25 mg. ones) were white.

Andrea was born, an apparently healthy child, in 1953. It was not until the 1970's when fears began to be publicly expressed that DES might be causing cancer in women exposed to it in utero that Andrea learned that she had been exposed to it while in her mother's womb. She was examined by a physician for potential DES-related problems and it was discovered that she had certain vaginal abnormalities, one called adenosis and the other a cervical hood, which had been identified as possible precursors to cervical cancer in DES-exposed women. Andrea then began to see a gynecologist regularly so that he could monitor the condition of her reproductive organs and examine any changes for signs of cancer.

In 1976, Andrea became pregnant in her left fallopian tube. This ectopic pregnancy made the removal of the tube necessary. While performing the operation, her physician discovered that Andrea's remaining fallopian tube and her uterus were deformed. The uterus in particular was very small and oddly shaped. Her physician felt these problems were also related to DES. Desiring to have her own children, Andrea also began seeing a physician specializing in fertility problems. Many tests, internal pelvic examinations, X-rays and an attempt to stretch her uterus followed. In 1978, Andrea had a normal uterine pregnancy, but the pregnancy was aborted within a few weeks due to a miscarriage. More tests, examinations and X-rays followed. In 1980, she became pregnant again, this time in her remaining right fallopian tube. Despite her doctors' efforts to save it, the tube had to be removed. Both tubal pregnancies caused severe pain. After the second ectopic pregnancy, Andrea became infertile. Her doctor referred her to an in vitro fertilization clinic but said there was very little likelihood her malformed uterus could carry a child to term even if she were to become pregnant by in vitro fertilization.

The case was submitted to the jury as a negligence claim.2 The issues below were whether Lilly was negligent in marketing DES or in failing to have an adequate warning accompany the drug, whether the white 25 mg. DES tablets were manufactured by Lilly and whether they were the cause or a significant contributing cause of plaintiff's two ectopic pregnancies, uterine miscarriage, malformed uterus and adenosis and cervical hood. The trial court made an unchallenged finding that, because of insufficiency of evidence, the jury could not find that the red 5 mg. DES tablets plaintiff's mother took while she was pregnant with plaintiff were manufactured by Lilly. Liability, therefore, could only be based on the 25 mg. pills. The case went to the jury in the form of special interrogatories pursuant to Federal Rule of Civil Procedure 49(a).

The jury found three injuries, causation and negligence. It awarded damages in the amount of $50,000. The special interrogatories and answers thereto are as follows:

A. Claim of Andrea Goldstein.

1. Was the DES taken by the plaintiff's mother in 1953 probably a cause of

a. The plaintiff's left ectopic pregnancy in 1976?

b. The plaintiff's right ectopic pregnancy in 1980?

c. The miscarriage of the plaintiff's uterine pregnancy in 1978?

d. Malformation of the plaintiff's uterus sufficient to interfere with normally carrying a pregnancy to term?

e. Plaintiff's adenosis and cervical hood?

If the jury has answered "No" to all of the above, that is the end of the case. If the jury has answered "Yes" to any of the above, proceed to question 2.

2. With respect to any injury to the plaintiff found by the jury to have been caused by her mother's use of DES, is it probable that white 25 mg. DES pills were a contributing cause of such injury?

If the answer to question 2 is "No", that is the end of the case; if "Yes", proceed to question 3.

3. Is it probable that the white DES pills found by the jury to have been a contributing cause of the plaintiff's injuries were manufactured by the defendant Eli Lilly & Co.?

If the answer to question 3 is "No", that is the end of the case; if "Yes", proceed to questions 4, 5 and 6.

4. Has the plaintiff proved by a preponderance of the evidence that the defendant was negligent in 1953 in distributing DES as a drug to be used in treating certain accidents of pregnancy?

If the answer is "Yes", proceed directly to question 6. If the answer is "No", proceed to question 5.

5. Has the plaintiff proved by a preponderance of the evidence that the defendant negligently failed to give adequate warning to prescribing doctors of the potential effect of DES on the unborn children of mothers who used it, which if given, would probably have prevented the injuries to the plaintiff?

6. What damages do you assess for the injuries claimed by the plaintiff which you find were caused by white DES pills manufactured by Eli Lilly & Co.?

$50,000

The defendants moved for judgment n.o.v. and alternatively for a new trial. The district court determined the jury finding that the 1976 ectopic pregnancy was caused by DES (interrogatory 1(a) ) to be against the clear weight of the evidence, and informed plaintiff that unless she agreed to strike the finding he would grant the defendant's motion for a new trial. Plaintiff agreed under protest and judgment was entered for plaintiff in the full amount of the damages found by the jury.

Defendant appeals the denial of its motions for directed verdict, judgment n.o.v. or a new trial. It claims that it is entitled to a new trial because the jury finding that DES caused the 1976 ectopic pregnancy, struck as erroneous, infected the whole verdict, and that striking it and entering judgment for the entire amount of damages found by the jury did not cure the error but compounded it. Defendant also claims grounds for a new trial in purported inconsistencies in the jury's answers to interrogatories and in errors in the trial court's charge to the jury. Defendant asserts that there should be reversal and judgment entered for it because there was no expert testimony that the 25 mg. DES tablets were the probable cause of any injury. It challenges the qualifications of two of plaintiff's expert witnesses and also asserts that their testimony only amounted to an opinion that there was a "possibility" DES caused the injuries found by the jury, not a "probability" as required by Massachusetts law. Defendant also asks for reversal and judgment in its favor on the grounds that there was not enough evidence in the record for the jury to find that the 25 mg. DES was made by Lilly. Defendant's various challenges to the judgment will be addressed in turn.

III. THE SETTING ASIDE OF THE JURY FINDING THAT DES CAUSED THE LEFT ECTOPIC PREGNANCY

The basic question arising from the striking of the jury finding but entering judgment on the full amount of damages is whether a new trial is mandated. The reason the district court gave for upholding the jury verdict on damages despite its striking of the jury finding on the left ectopic pregnancy was that the remaining injuries found by the jury to be caused by DES were adequate to "support the relatively modest jury award in this case."

Defendant argues that because there was no way for the trial court to determine what part of the jury damages award was allocated to the stricken injury finding, the court could not determine itself a proper award for the remaining injuries. This, defendant claims, invaded the province of the jury and infringed defendant's seventh amendment rights. In effect, defendant's position is that the district court's action can be viewed as either a factual determination and ruling that none of the jury award was assigned to the left ectopic pregnancy, or it can be viewed as something similar to an additur, reassigning the damages given for the stricken injury finding to the other injuries.

We think that a new trial is mandated. The defendant had the right to have the damages assessed by the jury. When the district court eliminated one of the injuries found by the jury it also eliminated one of the legs on which the damage award rested. If the amount of damages for the left ectopic pregnancy could have been determined accurately or closely approximated, then the remedy would have been a remittitur in that amount. Here, however, there was no way of allocating damages for each injury. The only remedy is a new trial. The Seventh Circuit has stated it as follows:

As a general proposition, subject to exceptions inapposite here, "[i]t is a deprivation of the right to a jury trial for the court to alter the verdict in matters of substance." Myrtle v. Checker Taxi Company, Inc., 279 F.2d 930, 934 (7th Cir.1960). Where, as here, there are multiple elements of compensatory damages put to the jury, and its verdict thereon is in general form, it is not proper for the court to attempt to segregate out elements which should not have gone to the jury by making its own assessment of the proven damages on the legitimate elements. Jayne v. Loder, 149 F. 21, 23 (3d Cir.1906). The problem, of course, is that the jury may or may not have reached the same conclusions as the court on the proper damage elements. It may in fact have relied wholly or substantially on what the court later decides to have been improper elements.

Durant v. Surety Homes Corp., 582 F.2d 1081, 1085-86 (7th Cir.1978) (footnotes omitted). See Arnold v. Eastern Air Lines, Inc., 681 F.2d 186, 206 (4th Cir.1982), cert. denied, sub nom. Aetna Casualty & Surety Co. v. United States, 460 U.S. 1102, 103 S. Ct. 1801, 76 L. Ed. 2d 366 (1983); Jacoby v. Johnson, 120 F. 487, 488 (3d Cir.1903).

If the court's action is viewed as an additur, it is constitutionally proscribed. Dimick v. Schiedt, 293 U.S. 474, 486-87, 55 S. Ct. 296, 301, 79 L. Ed. 603 (1935).

Plaintiff offers three arguments in defense of the judgment. The first is that whatever the errors flowing from the action of the district court's striking of special interrogatory 1(a) and entry of judgment on the original verdict they are harmless because the jury's answer to the interrogatory was not against the weight of the evidence and the judge erred in forcing the plaintiff to choose between agreeing to strike the jury finding or face a new trial. What the plaintiff seeks to do is challenge the trial court's striking of the jury finding.

There is a serious question as to whether plaintiff can now raise the issue that the struck answer was not against the weight of the evidence. Plaintiff asserts she should be able to argue this issue in defense of her judgment even though she agreed to strike the disputed finding and no cross appeal was taken.3 She relies for support on cases where a verdict winner who had accepted a remittitur was permitted to argue in defense of the remitted judgment, without a cross appeal, that the trial judge erred in requiring a remittitur in the first place.

We do not decide the difficult issue of whether the plaintiff can now challenge the trial court's conditional new trial ruling. Instead, we move ahead to the substantive argument that plaintiff makes--that the district court abused its discretion in striking the jury answer and ordering a new trial--and we determine that it did not.

It is settled law that a trial court's determination to set aside the verdict and grant a new trial is a matter of the exercise of the trial court's discretion and is not reviewable except for abuse of that discretion, lack of power, or where the court acts under the compulsion of a mistake of law. Allied Chemical Corp. v. Daiflon, Inc., 449 U.S. 33, 36, 101 S. Ct. 188, 190, 66 L. Ed. 2d 193 (1980); Valm v. Hercules Fish Products, Inc., 701 F.2d 235, 237 (1st Cir.1983). See 6A J. Moore, Moore's Federal Practice p 59.05 (2d ed. 1985) and cases cited therein; id. p 59.08, at 59-140 to -141, 151 to -152. The trial court may set aside a verdict and order a new trial if, in its opinion, the verdict is against the clear weight of the evidence, is based upon evidence that is false, or resulted from some trial error and amounts to a clear miscarriage of justice. See Borras v. Sea-Land Service, Inc., 586 F.2d 881, 886-87 (1st Cir.1978); Aetna Casualty & Surety Co. v. Yeatts, 122 F.2d 350, 352 (4th Cir.1941). In the past, the district court's exercise of discretion was virtually unreviewable. See, e.g., Portman v. American Home Products Corp., 201 F.2d 847, 848 (2d Cir.1953) (L. Hand, J.). In this circuit, however, we follow a modern trend and "exercise a closer degree of scrutiny and supervision where the basis for the new trial is the trial court's evaluation of the weight of the evidence rather than the trial court's feeling that an undesirable or pernicious element, such as trial error or prejudicial statements, obtruded into the trial." Borras, 586 F.2d at 887 (citations omitted) (quoting Lind v. Schenley Industries, Inc., 278 F.2d 79, 90 (3d Cir.), cert. denied, 364 U.S. 835, 81 S. Ct. 58, 5 L. Ed. 2d 60 (1960) ). This is so because "the trial court's discretion, although great, must be exercised with due regard to the rights of both parties to have questions which are fairly open resolved finally by the jury at a single trial." Coffran v. Hitchcock Clinic, Inc., 683 F.2d 5, 6 (1st Cir.), cert. denied, 459 U.S. 1087, 103 S. Ct. 571, 74 L. Ed. 2d 933 (1982). See Borras, 586 F.2d at 887. The trial judge should not act merely as a "13th juror" and set aside a verdict simply because he would have reached a different result had he been the trier of fact. Id.; Rios v. Empresas Lineas Maritimas Argentinas, 575 F.2d 986, 990 (1st Cir.1978).

We have reviewed the record in this case and, although we find that the issue of causation of the left ectopic pregnancy is a close one, we cannot say the trial court abused its discretion in making a finding that the jury verdict was so contrary to the weight of the evidence that it should be set aside and a new trial granted. Furthermore, the trial court stated a second ground for its conditional new trial order. The court found that confusion surrounding the admission of certain evidence on the issue of the causation of the left ectopic pregnancy amounted to "a miscarriage of justice, a gross error." The scope of our review of this basis for a new trial order is extremely limited. See Borras, 586 F.2d at 887.

The district court felt that through its fault an important piece of evidence was not adequately presented to the jury. The evidence involved the reading and interpretation of a report by two expert pathologists relative to a diagnosis of the condition of both fallopian tubes of plaintiff. The court felt that the way it had read and explained the report to the jury was misleading. The district court made the following remarks to plaintiff's counsel at the post-trial hearing on motions:

It should have been [in evidence]. I made a mistake, and you [plaintiff's counsel] talked me into it, so I consider that at least to that degree there was a miscarriage of justice, a gross error.

....

If I had put it to the jury properly, as [defendant's counsel] asked me to and you talked me out of doing, I would have put it to the jury that Dr. Driscoll agreed that the left ectopic--the left fallopian tube showed clear cell--follicular salpingitis.

In light of this mea culpa by the district judge, it would be presumptuous of us to find that his striking of the jury finding was an abuse of discretion. Our holding that the district court did not abuse its discretion in striking the jury finding makes it unnecessary, therefore, to decide whether plaintiff had a right to raise this issue.

Plaintiff's second argument in defense of her judgment is that under Rule 49(a) of the Federal Rules of Civil Procedure the issue of how much of the damage verdict the jury allotted to each finding of injury was an "omitted issue of fact." Under Rule 49(a), a party waives his right to a trial by jury on any issue of fact not specifically addressed by the jury in a special verdict "unless before the jury retires he demands its submission to the jury.... without such demand the court may make a finding; or, if it fails to do so, shall be deemed to have made a finding in accord with the judgment on the special verdict."

The rule has been uniformly upheld and followed. The problems that have arisen have not been over the rule's meaning, but concern its application and consequences. See Goeken v. Kay, 751 F.2d 469, 472 (1st Cir.1985); Fortier v. Dona Anna Plaza Partners, 747 F.2d 1324, 1336-37 (10th Cir.1984); Bell v. City of Milwaukee, 746 F.2d 1205, 1254 n. 62 (7th Cir.1984); Graphic Products Distributors, Inc. v. Itek Corp., 717 F.2d 1560, 1569 (11th Cir.1983); Wood v. Old Security Life Insurance Company, 643 F.2d 1209, 1215-16 (5th Cir.1981). Guidry v. Kem Manufacturing Co., 598 F.2d 402, 407 (5th Cir.1979), cert. denied, 445 U.S. 929, 100 S. Ct. 1318, 63 L. Ed. 2d 763 (1980).

We may assume arguendo with plaintiff that defendant's failure to request that the damages be allocated to each of the injuries on which liability was found triggered the strictures of Rule 49(a). This, however, does not end the matter. The rule gives the district court the authority to make a finding on the omitted question of fact; it does not give it the right to substitute its judgment for that of the jury on the question of damages. To the extent that the district court's finding is viewed as a Rule 49(a) finding, it must have determined that none of the $50,000 total damages were allocated by the jury to the left ectopic pregnancy. In light of the evidence and the jury's other findings, this was "clearly erroneous."

We have read the record carefully relative to the left ectopic pregnancy and it discloses the following. There was testimony that the left ectopic pregnancy inflicted great pain and suffering on the plaintiff, that it caused her to undergo an operation and to incur considerable medical expenses, and contributed to her infertility. There was no testimony that the malformed uterus caused any direct pain, and it was not found that it prevented uterine pregnancies, only interfered with them. This distinction was made clear to the jury by the district court in answer to a question from the foreman during the jury's deliberations. Considering the often central role testimony about the ectopic pregnancies played in the trial, it must be assumed that the jury award reflected to some degree the trauma, pain and suffering, medical expenses and consequences of this injury. Even if the damages award in this case is viewed as surprisingly low, we cannot escape the conclusion that it would have been even lower if the jury had not found the left ectopic pregnancy to have been caused by DES.

Plaintiff's third argument is that the judge's attempt to cure the jury's erroneous finding of injury is not fatal to the verdict and a new trial is not required. She argues that in light of the evidence of considerable physical damage to the plaintiff, large medical bills, permanent infertility, pain and suffering, and fear of cancer the jury verdict was so low that the district court could decide that an error in one of the jury's findings of injury was harmless. In effect, plaintiffs ask us to determine that the jury would have made the same $50,000 damage verdict even if it had not found DES caused the left ectopic pregnancy. This, of course, is what the judge did. This argument misses the point. The defendant was entitled to have the jury, not the judge, determine damages. After the district court determined that one of the liability findings was against the weight of the evidence, it had two choices: it could get plaintiff to agree to a remittitur, if there was an evidentiary basis for one, or it could order a new trial. It could not strike one of the grounds for the jury's award of damages and accept the award in toto.

Plaintiff argues that if a new trial is granted it should be restricted to the issue of damages since that is "commonly" done in cases like this. The Supreme Court has stated the standard for granting a partial new trial in Gasoline Products Co., Inc. v. Champlin Refining Co., 283 U.S. 494, 51 S. Ct. 513, 75 L. Ed. 1188 (1931). It held that an issue may not be separated for retrial "unless it clearly appears that the issue to be retried is so distinct and separable from the others that a trial of it alone may be had without injustice." Id. at 500, 51 S.Ct. at 515. Gasoline Products involved a contract action where in order for the jury at retrial to determine damages it would have also had to determine material facts not yet established that bore on liability under the contract. The Court held that the "question of damages ... is so interwoven with that of liability that the former cannot be submitted to the jury independently of the latter without confusion and uncertainty, which would amount to a denial of a fair trial." Id.

The issue of damages here is so inextricably interwoven with that of liability as to require a new trial on all issues. Liability and damages cannot be separated because damages can only be found by assessing the same evidence as to the nature, extent and results of the injuries that the jury must consider to determine liability. Moreover, the plaintiff's fertility problems were all treated together and were thought by her doctors to be caused by DES. Evidence relating to virtually all of the plaintiff's gynecological examinations, operations and treatment would necessarily be introduced at a new trial limited to a determination of damages for the malformed uterus and adenosis and cervical hood. This would cause confusion and uncertainty and result in prejudice amounting to a denial of a fair trial.

Since there has to be a new trial on all issues because of the striking of the jury answer, we need not consider whether inconsistencies in the jury's answers and/or errors in the charge also require a new trial.

V. THE SUFFICIENCY OF PLAINTIFF'S EVIDENCE ON THE ISSUE OF CAUSATION

A. Admissibility of the Testimony of Doctors Hagen and Gross

Plaintiff had to prove that ingestion of the 25 mg. pills from the fifteenth through the thirty-fifth weeks of her mother's pregnancy was probably a substantial contributing factor to her injuries. Defendant challenges the district court's decision to admit the testimony of two of plaintiff's expert witnesses, Doctors Hagen and Gross, on this causation question. It asserts that the two doctors are clinicians and not research scientists and that without specialized knowledge of the research-type causation question at issue neither should have been permitted to testify. Furthermore, the defendant claims that neither doctor was able to point to any research, investigation or any other evidence which supported his opinion that plaintiff's injuries were causually related to her mother's ingestion of the 25 mg. DES rather than the earlier 5 mg. pills.

It is settled law in this circuit that "[w]hether a witness is qualified to express an expert opinion is a matter left to the sound discretion of the trial judge. In the absence of clear error, as a matter of law, the trial judge's decision will not be reversed." A. Belanger & Sons, Inc. v. United States, 275 F.2d 372, 376 (1st Cir.1960); McDonald v. Federal Laboratories, Inc., 724 F.2d 243, 248 (1st Cir.1984); Garbincius v. Boston Edison Co., 621 F.2d 1171, 1174 (1st Cir.1980). Defendant fails to convince us that we should reverse the district court's decision to admit the expert testimony of these two doctors.

Defendant is correct that under Federal Rule of Evidence 702, an expert must possess some specialized knowledge of the field in which he purports to be an expert, but defendant is incorrect insofar as it challenges the experts here for not having specialized knowledge as to the specific question at issue. See C. McCormick, Evidence Sec. 13, at 30 & n. 70 (E. Cleary ed. 1972). We have said that a physician is qualified to give an opinion as to the mental health of someone even if he is not a psychiatrist. Alvarado v. Weinberger, 511 F.2d 1046, 1049 (1st Cir.1975). The fact that the physician is not a specialist in the field in which he is giving his opinion affects not the admissibility of his opinion but the weight the jury may place on it. Alvarado, 511 F.2d at 1049. Under Rule 702,

[i]f scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.

There is no dispute here that Doctors Hagen and Gross were qualified as experts in the field of medicine and that teratology, the study of abnormal development and congenital malformations, is scientific knowledge that is part of the field of medicine. The doctors are board-certified obstetrician-gynecologists who are also clinical instructors at Harvard Medical School. They gave their opinions on the teratologic effects of DES. The district court did not abuse its discretion in permitting them to do so.

Defendant also claims, under Rule 703, that the plaintiff's experts could point to no "facts or data ... of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject." The record belies this assertion. There was testimony at trial that the teratologic effects of drugs on unborn children was a matter taken up in the general medical education of both doctors and in their specialized training in obstetrics and gynecology. Although neither doctor had done systematic research on the matter, both had studied some of the articles published regarding DES and embryology, and Dr. Gross had attended lectures and engaged in discussions with researchers in the field. Furthermore, Dr. Gross had from 60 to 80, and Dr. Hagen from 100 to 150 DES-exposed women as patients. Not only did both have considerable clinical experience with DES-related problems but the number of DES daughters in their practices led them to seek current information on the effects of DES on unborn children. Both doctors also treated and observed the plaintiff over several years. The fact that defendant was able to undercut some of the research basis for the doctors' opinions does not affect the admissibility of those opinions. If the factual underpinnings of their opinions were in fact weak, that was a matter affecting the weight and credibility of their testimony. Coleman v. DeMinico, 730 F.2d 42, 47 (1st Cir.1984).

Defendant asserts that it is entitled to judgment because the testimony offered by plaintiff's experts was legally insufficient to support the jury finding in response to interrogatory 3 that it was "probable that the white 25 mg. DES pills were a contributing cause" of plaintiff's injuries. Defendant reads the expert testimony of the plaintiff's witnesses to amount to no more than a collective opinion that there was a possibility that the 25 mg. DES was a cause or contributing cause of the injuries. Under Massachusetts law, applicable in this diversity action, the plaintiff must prove it is "more likely than not that the conduct of the defendant was a substantial factor in bringing about the harm." Restatement (Second) Torts Sec. 433B(1) comment a (1965); LaClair v. Silberline Manufacturing Co., 379 Mass. 21, 393 N.E.2d 867, 873 (1979) and cases cited therein. An opinion that a defendant's conduct is only a "possible cause" is insufficient to establish causation under Massachusetts law. Id.; Swartz v. General Motors Corp., 375 Mass. 628, 378 N.E.2d 61, 64-65 (1978). "A verdict may not be based on conjecture and surmise, and expert opinion does not help if it is demonstrated that it rests on speculation." Id. 378 N.E.2d at 65. Defendant does not challenge the standard set forth in interrogatory 3 for the determination of proximate causation. Nor does defendant challenge the finding of the jury that the plaintiff's malformed uterus and adenosis and cervical hood were caused by her in utero exposure to DES. Rather, defendant challenges the sufficiency of the evidence to support a finding that the 25 mg. DES pills as opposed to the 5 mg. pills were probably a contributing cause of plaintiff's injuries.

When considering an appeal from the denial of motions for directed verdict and judgment n.o.v., this court must view all the evidence in the light most favorable to the plaintiff, and draw all reasonable inferences in the plaintiff's favor. To overrule the denial of defendant's motions we must find that the facts and the inferences drawn from the facts can reasonably lead to only one conclusion. Hubbard v. Faros Fisheries, Inc., 626 F.2d 196, 199-200 (1st Cir.1980); Rios v. Empresas Lineas Maritimas Argentinas, 575 F.2d 986, 989 (1st Cir.1978). The expert testimony in this case was certainly not free of ambiguity and uncertainty. But it is a matter for the jury to resolve any inconsistencies in expert testimony. CVD, Inc. v. Raytheon Co., 769 F.2d 842, 849 (1st Cir.1985); Sears, Roebuck & Co. v. Penn Central Company, 420 F.2d 560, 564 & n. 7 (1st Cir.1970); Calderone v. Wright, 360 Mass. 174, 274 N.E.2d 588, 589 (1971).

On our reading of the record, we determine that the jury had evidence from which to find the 25 mg. DES was probably a contributing cause of the plaintiff's injuries. There was uncontradicted testimony that the DES-affected organs developed from a common embryonic unit, called the mullerian duct, from the sixth through the twenty-second weeks of pregnancy. The testimony was that the DES caused the malformations sometime during this period of development. There was also testimony that the amount and duration of DES exposure have a significant effect on the development of the reproductive organs. The 5 mg. DES was taken from the ninth through the fourteenth weeks of pregnancy, and the 25 mg. DES was taken from the fifteenth week on. During the period of development of the reproductive organs, the 25 mg. DES was ingested in significantly higher amounts and for two weeks longer than the 5 mg. DES. It is a reasonable inference from these facts that the 25 mg. DES probably contributed significantly to the plaintiff's injuries.

Defendant makes much of the fact that there was testimony that the 5 mg. DES taken alone would probably have caused the injuries, and argues that testimony that the 25 mg. DES taken alone would have caused the injuries was irrelevant since the 25 mg. DES was not taken until after the 5 mg. DES had been ingested. But, there was also testimony that, "[i]n terms of probability, these abnormalities could have occurred from the five-milligram pills, and they could have occurred from the 25-mg. pills, and it is impossible to separate them." (Emphasis added.) Defendant is correct that the plaintiff's expert witnesses did not know what the relative causal roles of the 5 and 25 mg. DES dosages were and did not testify that the 25 mg. DES "probably" caused a definite amount of damage. Under the law of Massachusetts, however, where there is evidence that two causes for which liability may attach are probably involved in an injury but their respective causal roles cannot be separated, joint and several liability may attach to each cause. Chase v. Roy, 363 Mass. 402, 294 N.E.2d 336, 340 (1973); Feneff v. Boston & Maine Railway Co., 196 Mass. 575, 82 N.E. 705, 707 (1907); Shultz v. Old Colony Street Railway, 193 Mass. 309, 79 N.E. 873, 877 (1907); Oulighan v. Butler, 189 Mass. 287, 75 N.E. 726, 728 (1905); Corey v. Havener, 182 Mass. 250, 65 N.E. 69, 69 (1902); Boston & Albany Railroad Co. v. Shanly, 107 Mass. 568, 579 (1871); see also Shepard v. General Motors Corporation, 423 F.2d 406, 408 n. 2 (1st Cir.1970); Whalen v. Shivek, 326 Mass. 142, 93 N.E.2d 393, 397 (1950); cf., In re Sevigny's Case, 337 Mass. 747, 151 N.E.2d 258 (1958) (where the underlying causes of a disease were entirely unknown, one expert's opinion that it was more than possible that an infection of the finger caused the deceased's leukemia was not sufficient to support a jury finding of liability but was mere speculation). Here, there was sufficient evidence for the jury to find that both dosages probably contributed to the malformations that occurred sometime during the organs' development. The fact that the experts could not separate the particular effects of each dosage did not preclude the jury from finding defendant liable under Massachusetts law.

VI. THE PROOF THAT DEFENDANT MANUFACTURED THE 25 MG. DES TAKEN BY PLAINTIFF'S MOTHER

Defendant asserts that plaintiff failed to produce enough evidence to permit the jury to find that the 25 mg. DES that plaintiff's mother ingested was manufactured by defendant. Plaintiff's proof consisted of a chain of inferences in which every link was necessary for the jury to find that defendant manufactured the 25 mg. DES at issue. The necessary inferences the jury had to draw from the testimony before it were: (1) that plaintiff's mother was given a generic prescription for DES; (2) that this generic prescription was filled at the Trachtenberg & Meyers Pharmacy; (3) that because Trachtenberg & Meyers did not stock 25 mg. DES the pharmacist ordered the product from its supplier Gilman Brothers; (4) that the order was placed with Gilman Brothers as an "open order," one which did not specify any brand of DES; and (5) that Gilman Brothers had a policy to fill open orders with defendant's products.

Defendant contends that the fourth inference has no evidentiary foundation. It claims that the evidence did not show that it was more likely that, when presented with plaintiff's mother's generic prescription, the pharmacy would have placed an open order rather than requesting the cheapest brand or specifying a particular brand of the drug from Gilman.

The testimony supporting the inference was as follows. The pharmacy had a policy not to keep supplies of 25 mg. DES in stock because there was little call for it. It also had a policy to order only drugs manufactured by Lilly or the Berkeley Drug Company unless it was presented with a prescription specifying the brand-name drug of another manufacturer. When presented with plaintiff's mother's generic prescription for 25 mg. DES, then, the pharmacy would have ordered just enough of it from Gilman to fill the prescription. It could have, however, specified Berkeley or placed an open order and received Lilly. Defendant argues that it is more likely that the pharmacy ordered Berkeley DES from Gilman because Berkeley was over 40% cheaper than Lilly. Although it is true that Berkeley drugs were significantly cheaper than Lilly drugs, they only came in bottles of 1,000, whereas Lilly's drugs came in bottles of 100. For the pharmacy to order Berkeley's DES so as to fill the 25 mg. DES prescription for 475 pills, it would have had to purchase more than twice the quantity needed. This would have been contrary to pharmacy policy and uneconomical because it would have left the pharmacy with 525 pills it might never sell. Moreover, if the pharmacy had ordered Berkeley DES from Gilman to fill the prescription, it could be inferred that the witness Meyers, who was a pharmacist at Trachtenberg & Meyers at that time, would have seen the half-empty bottle on the shelf and remembered it. Berkeley pills came in very large bottles and were stored in direct view of where the pharmacists prepared prescriptions. Meyers testified that he had never seen a bottle of 25 mg. Berkeley DES in the store. Based on this evidence, it could be inferred that the DES prescription was filled by placing an open order with Gilman who supplied DES manufactured by defendant.

Reversed and remanded for a new trial on all issues. No costs to either party.