Ohalos, Chapter 9, Mishnah 3. Our chapter runs a long series of fourteen mishnayos, all of them built around one object: the kaveres, a beehive standing in the doorway of a house, with its mouth facing outward toward the street. Everything turns on whether the tumah spreads up, down, inward or outward, and the guiding rule so far has been that a keli does not stop tumah; only an ohel stops tumah. A keli merely carries the tumah along with it. This is the third mishnah in the series, and it takes that whole discussion and sets a boundary around it.
The mishnah opens by limiting what came before: "Bameh devarim amurim?" In what case was all of that said? "Bizman shehi keli mechulchal." Everything we have learned until now, or at least most of it, applies when the kaveres is a genuine keli that is mechulchal, that is, a working beehive with openings in it, and those openings were stuffed loosely with straw. A beehive by its nature needs holes so that the bees can come and go. As long as it is that kind of vessel, plugging the holes with straw has only a limited effect, and the object remains a keli.
Two Cases Where It Is No Longer a Keli
"Haysah pechusa ufekukah bekash." The hive itself split open, and he packed the break with straw. According to the Bartenura, the mishnah here has in mind a kaveres that was never serving as a home for bees at all: there are no bee openings in it whatsoever. What happened was that it cracked, and cracked so badly that it can no longer function as a vessel, and he filled the fissure with straw.
"O afutza." That is the second case: this one truly is a working hive, and it truly does have openings, since bees need a way in and out, but rather than pushing straw in loosely he closed those openings up tightly. And the mishnah supplies the definition by way of a question: "Eizu hi afutza?" What earns a hive that name? Any hive in which there is not a tefach of open space at any single point, because he blocked it firmly and completely. (There are those who understand this case too as dealing with a hive that has broken.) Whichever the reason, the object in front of us now is no longer a keli. And that single fact overturns everything, for it now carries the status of an ohel, and an ohel does hold tumah back.
The Four Cases of the Mishnah
Tumah underneath. "Kezayis min hameis nasun tachtav, kenegdo ad hatehom tamei." A kezayis of a corpse is lying below the hive. Whatever is straight beneath it is tamei, as far down as one can reach, and nothing more. What of an object resting directly above it on the same vertical line? Tahor. Since we are dealing with something that is not a keli but an ohel, it forms a barrier. The tumah sinks to the depths, yet it cannot climb past the kaveres.
Tumah on top. "Al gabav, kenegdo ad harakiya tamei." As in the earlier mishnayos of the chapter, the area under discussion is the space outside the house. When the kezayis of the corpse is placed on the upper surface of the hive, whatever is straight above it is tamei to the very heavens, since the tumah rises with nothing to obstruct it. Everything else, though, remains tahor: the inside of the hive is tahor, and so is the area under it. Why not rule, as we would have ruled in the earlier cases, that the tumah penetrates downward through the hive? Because we have no keli here. Not being a keli, it carries the din of an ohel, and an ohel brings the tumah to a halt.
Tumah in the house. "Vahabayis, ein tamei ela habayis." When the kezayis lies somewhere within the house, it is the house that becomes tamei, while the hive's own space does not. All the rest is tahor, since piv lachutz, the mouth of the hive faces the street. By what route would tumah get inside? The kaveres constitutes an independent ohel, and nothing penetrates into it.
Tumah inside the kaveres. "Besocho, ein tamei ela socho." Should the kezayis be lying within the hive, only its inner space is tamei. There is nowhere for the tumah to go: no opening measures a tefach, and the bee openings were not merely filled with loose straw but shut tight.
Comparing This to the Earlier Cases
It is worth setting these four rulings beside the halachos that came before them. So long as the kaveres was a functioning vessel, tumah resting above it penetrated downward, rendering a keli that lay under the hive tamei, and tumah lying below it rose upward, rendering a keli on top of the hive tamei. Almost all of those rulings flowed from one rule: keilim erect no barrier against tumah. To this the mishnah now attaches its qualification: bameh devarim amurim? Only where we are dealing with a keli. A beehive has to be serving as a beehive, with the openings and windows through which the bees pass. Where nothing was done beyond pushing straw into those openings, the effect is only partial and the object retains its status as a keli.
But if it is no longer a keli, whether because it broke (and even though the broken places were filled with straw, it is not a functioning beehive, since it has no holes at all), or because it does have holes but they were completely sealed, or sealed down to less than a tefach, then the calculation is reversed.
Take, then, tumah lying underneath. We no longer say that it penetrates through to the upper side. Downward it travels with no limit at all, reaching the tehom, but the instant it rises and encounters the hive its progress ends. Being an ohel rather than a keli, this hive forms a barrier against tumah. With tumah resting on top, it ascends without limit, nothing standing in its path, straight up to the rakia, yet a keli sitting below the hive is unaffected. It is keilim alone that fail to hold tumah back; an ohel holds it back. In the earlier halachos, where we were dealing with a keli, tumah on the upper surface penetrated through, and even in a version where the vessels within were shielded, the vessels below contracted tumah.
As for the third case, that tumah in the house renders only the house tamei, this was in truth the halachah in the earlier cases as well; the mishnah simply states it for consistency.
The fourth case is where the contrast stands out most sharply. Until this point, with tumah lying within a hive that counted as a keli, we had said either that the tumah made its way out through the bee openings and rendered the entire house tamei, or that vessels above and below contract tumah because we treat the hive's whole airspace as though tumah filled it. Yet all of that rested purely on its being a keli, and a keli forms no barrier. Whether we picture the vessel as packed with tumah, so that the tumah automatically travels up and down, or we picture it escaping by way of the openings and spreading in the house, in neither version does the vessel obstruct anything. Here, however, there is no keli, and therefore tumah placed within stays within. An ohel keeps the tumah contained, and there exists no petach of a tefach by a tefach through which it might leave. The inside is tamei, and everything beyond it is tahor.