MISHNA: After conducting the search, one need not be concerned that perhaps a marten dragged leaven from house to house, or from place to place, placing leaven in a house that was already searched. As if so, one need also be concerned that perhaps leaven might have been dragged from courtyard to courtyard and from city to city. In that case, there is no end to the matter, and it would be impossible to rely on any search for leaven.

GEMARA: The Gemara infers from the mishna: The reason that one need not search again is that we did not see the marten drag the leaven from the house; however, if we saw the marten drag leaven from the house, we are indeed concerned that it dragged the leaven into the second house, and it therefore requires searching for leaven.

The Gemara raises a difficulty: But why is this so? Let us say that the marten ate the bread it took. Didn’t we learn in a mishna: The residences of gentiles are ritually impure, as their wives may have miscarried, and due to the fact that gentiles would bury their stillborn babies in their houses, all their residences are deemed ritually impure due to the possibility of impurity imparted by a corpse. And how long must a gentile have stayed in a residence for the residence to require searching? He must have lived there for forty days. The reason is that until forty days after conception the miscarried fetus is not classified as a stillborn, as it is not sufficiently developed before that stage.

This mishna continues: And this decree applies even though the gentile resident has no wife. In issuing the decree, the Sages did not distinguish between a married couple and a single man, so that people would not err in its application (Me’iri). And any place where a marten or a pig can enter unimpeded need not be searched, as presumably if a stillborn was buried there, one of these animals would have taken it. As this mishna indicates that there is a presumption that martens eat whatever they find, the Gemara suggests that the same should apply to leaven. Therefore, even if one actually saw the marten take the bread, he can assume that the animal ate it, obviating the need for an additional search.

Rabbi Zeira said: This is not difficult, as in this case, where no search is required, it is referring to flesh, whereas in that case, where one is required to search again, it is referring to bread. Rabbi Zeira elaborates: With regard to flesh, a marten does not leave remnants behind, and therefore the stillborn would have been entirely consumed. With regard to bread, however, the marten leaves remnants behind, requiring an additional search.

Rava said: What is this comparison? These cases are not comparable. Granted, there, with regard to the stillborn, one could say that it was in the house and one could say that it was not in the house. And even if you say it was there, say that the marten ate it. The very presence of the stillborn in the house is based on an assumption, and even if it was there, it was probably consumed. However, here, where one definitely saw the marten take the bread, who will say that the marten ate it? It is a conflict between an uncertainty whether or not the marten ate the bread, and a certainty that the bread was there. The principle is that an uncertainty does not override a certainty.

The Gemara raises a difficulty against this principle: And is it so that an uncertainty does not override a certainty? But wasn’t it taught in a baraita: With regard to a ḥaver who died and left a storehouse filled with produce, even if the produce was there only that day, the fruit has the presumptive status of produce that was ritually prepared, i.e., properly tithed. The assumption is that the owner tithed it himself or commanded others to do so. However, here, this produce was certainly untithed at the outset, and there is uncertainty whether they are tithed or whether they are not tithed. Despite this conflict, the uncertainty whether they were tithed comes and overrides the certainty that they were untithed produce.

The Gemara rejects this contention: There, the conflict is between certainty and certainty, as the produce is certainly tithed, in accordance with the statement of Rabbi Ḥanina Ḥoza’a. As Rabbi Ḥanina Ḥoza’a said: There is a presumption with regard to a ḥaver that he does not allow produce that is not ritually prepared, to leave his possession. It is therefore certain that the produce is tithed.

And if you wish, say instead that in that case the conflict is between uncertainty and uncertainty, as perhaps one could say that the produce was never initially untithed. It is possible that there was never an obligation to tithe this produce, in accordance with the opinion of Rabbi Oshaya.

As Rabbi Oshaya said: A person may employ artifice to circumvent obligations incumbent upon him in dealing with his grain, and bring it into the courtyard in its chaff so that his animal may eat from it. And this grain is exempt from tithes. Although the obligation to tithe produce that has been fully processed applies even to animal fodder, it is permitted to feed one’s animal untithed produce that has not been fully processed. In light of this halakha it is possible that the fruit of the ḥaver in the storehouse may not have been subject to the obligation to be tithed. Consequently, the aforementioned case involving produce is a conflict between two uncertain factors, as it is uncertain whether or not the owner was obligated to tithe the produce in the first place, and even if he was required to do so, it is uncertain whether or not the ḥaver tithed it.

The Gemara raises a further difficulty against Rava: And is it so that an uncertainty does not override a certainty? But wasn’t it taught in a baraita that Rabbi Yehuda said: There was an incident involving the maidservant of a certain violent person in the city of Rimon, who threw a stillborn baby into a pit,