that anyone who is eligible for levirate marriage is eligible for ḥalitza and anyone who is not eligible for levirate marriage is not eligible for ḥalitza. Therefore, the original understanding of Rabbi Yoḥanan’s opinion, that both the intercourse and the ḥalitza of a pregnant woman are valid, was accurate.

Rava therefore provides a different defense of Rabbi Yoḥanan’s opinion: Rather, Rava said that this is what the baraita is saying: In the case of one who consummates the levirate marriage with his yevama under the assumption that there is a mitzva to do so, and then she is found to have been pregnant at the time of the intercourse, a rival wife of this yevama may not marry lest the offspring be viable, and intercourse with a woman pregnant with viable offspring is not considered a valid consummation of levirate marriage through intercourse, and ḥalitza of a woman pregnant with viable offspring is not considered effective ḥalitza. And furthermore, even if the offspring is viable, it does not release her and her rival wives from the levirate bond until it comes into the air of the world, i.e., until it is actually born.

It is taught in a baraita in accordance with the opinion of Rava: In the case of one who consummates a levirate marriage with his yevama, and then she is found to have been pregnant at the time of the intercourse, a rival wife of this yevama may not marry lest the offspring be viable. This is because intercourse or ḥalitza with a woman pregnant with viable offspring does not release a yevama from the levirate bond; rather, the offspring releases her. And furthermore, even if the offspring is viable, it does not release her and her rival wives from the levirate marriage bond until it comes into the air of the world.

The Gemara explains that the baraita appears to contradict Reish Lakish’s opinion: According to the baraita, the only reason that levirate marriage with the pregnant yevama does not permit the rival wife to marry is as the baraita stated: Lest the offspring be viable. By inference, were the offspring not viable, her rival wife would be released from the levirate bond. If so, let us say that this baraita is a conclusive refutation of the opinion of Reish Lakish.

The Gemara defends Reish Lakish’s opinion: Reish Lakish could have said to you: This is what the baraita is teaching: In the case of one who consummates a levirate marriage with his yevama, and then she was found to have been pregnant at the time of the intercourse, a rival wife of this yevama may not marry lest the offspring not be viable, which would mean that all the rival wives are bound by the levirate bond. And even if one of those wives has intercourse or performs ḥalitza with the yavam, it would be ineffective in releasing them from the levirate bond because ḥalitza with a pregnant woman is not considered effective ḥalitza, and intercourse with a pregnant woman is not considered a valid consummation of levirate marriage through intercourse.

Reish Lakish explains the need for the final clause of the baraita: And even if you say: Let the wives marry without the need for any levirate marriage or ḥalitza because one should follow the majority of women, and the majority of women give birth to a full-fledged, i.e., viable, offspring, and therefore one should presume no levirate bond exists, to counter this claim the baraita concludes: Even if the offspring will be viable, an offspring does not release a yevama and her rival wives from the levirate bond until it comes into the air of the world.

Rabbi Elazar said: Is it possible that there is halakhic acceptance of this opinion of Reish Lakish and it was not hinted to by something taught in the Mishna? He went out of the study hall, carefully checked the mishnayot, and found one that supported Reish Lakish’s opinion, as we learned in a mishna: In the case of a woman whose husband and rival wife went overseas, and then witnesses came and said to her: Your husband died, and her husband had a brother, this woman may neither marry someone other than his brother, nor may she enter into levirate marriage with that brother, until she knows whether perhaps her rival wife is pregnant. If she discovers that her rival wife is not pregnant, she would then be able to perform levirate marriage or ḥalitza. If she discovers her rival wife is pregnant, she would have to wait to see if the pregnancy is viable. If it is found to be viable, only at that point would she be permitted to marry someone else.

Rabbi Elazar explains how this mishna supports Reish Lakish’s opinion: Granted, she may not enter levirate marriage as perhaps her rival wife is pregnant and the offspring will be viable, and therefore by consummating the levirate marriage the yavam would encounter the Torah prohibition against engaging in relations with one’s brother’s wife. But why may she not perform ḥalitza? Granted, she may not perform ḥalitza during the first nine months following her husband’s death and then proceed to also marry during those nine months; this is prohibited due to the fact that there is the uncertainty whether her rival wife is pregnant with viable offspring, in which case she would be released from the levirate bond.

But let her perform ḥalitza during the first nine months following her husband’s death and then wait to marry until after those nine months. By that point in time, even if the rival wife was pregnant she would have already given birth. If the offspring was viable, then it emerges that there was never a levirate bond, and if it is not viable, then she was released from her levirate bond through the ḥalitza she performed. Either way, she would now be permitted to remarry. Why, then, does the mishna not consider this possibility? Rabbi Elazar claims that the only explanation for this is if the mishna assumes that ḥalitza performed while one of the wives of the deceased is pregnant is not effective. As such, the mishna is a proof for Reish Lakish’s opinion.

The Gemara rejects Rabbi Elazar’s proof: But even according to your reasoning that ḥalitza with a pregnant woman is not effective, the mishna should have considered an additional possibility: Let her perform ḥalitza and marry, doing both after nine months have passed since the death of her husband. Doing so should be effective according to both Reish Lakish and Rabbi Yoḥanan.

Rather, the discussion of this topic should be held apart from that mishna, as the true reason for that mishna’s ruling is as it is Abaye bar Abba and Rav Ḥinnana bar Abaye who both say that she is prohibited from performing ḥalitza while one of the wives of the deceased is pregnant because perhaps the offspring will be viable, in which case any ḥalitza performed would be entirely unnecessary and therefore meaningless, and she would remain permitted to marry into the priesthood, as the opening mishna of the chapter rules. However, in this situation people might not realize that the ḥalitza she performed was meaningless, and they would think she is a ḥalutza, who is prohibited from marrying a priest. And it would therefore emerge that if she were allowed to perform ḥalitza while pregnant, it is possible that you will ultimately require a public announcement to be made for her to attest to the fact that she is in fact still permitted to marry into the priesthood.

The Gemara wonders why this poses a problem: But why not let her perform ḥalitza while still pregnant, and then if it becomes necessary, require a public announcement to be made for her? The Gemara explains why one should avoid having to rely on a public announcement: Perhaps there were some people who were present at the ḥalitza and were not present at the public announcement, and when the courts permit her to marry a priest they might come to say that they are permitting a ḥalutza to marry a priest.

Abaye suggests another rejection of Rabbi Elazar’s proof from that mishna: Abaye said to Rabbi Elazar: The very formulation of the mishna refutes Reish Lakish’s opinion, as does the mishna teach that she may neither perform ḥalitza nor enter into levirate marriage? No, the mishna teaches only that she may neither be married nor enter into levirate marriage, which implies only that she may not marry without first performing ḥalitza, but if the yavam performs ḥalitza with her, she would indeed be permitted to marry after nine months have passed since her husband’s death. This understanding of the mishna undermines the basis of Rabbi Elazar’s proof from the mishna.

Even if that mishna does not support Reish Lakish’s opinion, nevertheless it is taught in a baraita in accordance with the opinion of Reish Lakish: In the case of one who performs ḥalitza with a pregnant woman and she miscarries, she requires another ḥalitza with the brothers in order to release her from the levirate bond. The baraita assumes that the original ḥalitza is ineffective because it was done while she was still pregnant, which is in accordance with the opinion of Reish Lakish.

Rava said: The halakha is in accordance with the opinion of Reish Lakish in these three disputes: One, this dispute that we already stated with regard to the ḥalitza of a pregnant woman. The other dispute concerns that which we learned in a mishna: In the case of one who verbally divides up his possessions among his descendants, stating how he wishes his estate to be divided after his death, if he increases the proportion of his estate that should go to one of his children or decreases the proportion of his estate that should go to another one of his children, or if he equally distributes between them the double portion of the firstborn, then his words are binding.

But if he said explicitly that the receipt of those portions should be considered as an inheritance, then it as though the verbal division of his property that he said is nothing, i.e., it is non-binding, since his words directly contradict the halakhot of inheritance as they are written in the Torah. However, if he wrote a will and somewhere therein he wrote, whether at the beginning, whether at the end, or whether in the middle, that the receipt of the portions should be considered as a gift, as opposed to an inheritance, then his words are binding.