Ohalos, Chapter 5, Mishnah 6. Our mishnah picks up exactly where the previous one left off, and it opens by demanding an explanation of what was just taught.
The earlier halachah established that a kli which cannot receive tumah, once it is spread over the opening joining a lower room to the story above it, leaves that upper story entirely tahor even while a corpse lies in the room beneath. Keilim that are tehorim halt tumah from rising. That teaching, however, ended with an essential condition: such a kli succeeds only "im dafnos ohalen," in partnership with the partitions of the ohel. Hence the present mishnah opens with "keitzad," meaning: describe for us precisely how this works, and it responds through a sequence of cases.
The bor and the dos with a chafishah spread over them
"Habor vehados shebabayis uchefishah nesunah alav." Both a bor and a dos are hollows carved out of the earth, dug primarily to store water. The bor is a plain excavated pit, while a dos has a lining built along its interior. Picture such a hollow inside a dwelling, with a chafishah laid across its mouth. The chafishah is an enormous basket whose capacity reaches forty se'ah of liquid, and a receptacle that big is entirely incapable of receiving tumah. Baskets of this sort served the olive harvest: olives were loaded in, they heated up and grew soft, and the pressing afterward went far more easily.
Now a meis is present in that dwelling. The mishnah rules that the contents of the bor, or of the dos, stay tahor.
The picture is worth redrawing. Earlier we spoke of a ground-floor room with a second story above it. Here the pit plays the role of the lower level, a kind of cellar, and the dwelling's floor is the level above. The corpse rests on that floor, and whatever sits down in the cellar is rescued.
What accounts for this? The contour of the pit's opening. Imagine crouching on the floor next to this bor or dos and releasing a small marble along the ground: the marble would not tumble in. That is because the mouth is not flush with the floor surface. A slight lip, a stub of partition, rises around it. So here you have a fragment of an ohel's partition, and resting upon it the huge basket, which is incapable of tumah on account of its volume. Partition and kli combine, and everything down in the pit escapes the tumah.
A smooth well and an uncovered kaveires
"Im hoysah be'er chalakah." A be'er is a well, and chalakah, smooth, indicates that its opening sits perfectly even with the floor. Nothing juts up around it. Release your marble and it falls straight down.
In the same breath the mishnah introduces a parallel case: "o kaveires." A kaveires is a specific style of basket manufactured to house bees. It will appear repeatedly throughout Masechtas Ohalos. It, too, cannot receive tumah. The mishnah calls it "pesuchah," open, indicating that it has not been stuffed with straw. (Some texts read "kefuyah," turned upside down, but the version before the Bartenura is "pesuchah.")
"Uchefishah nesunah aleha." That same giant olive basket now rests upon the mouth of the level well, or upon the open beehive. With the well, the well's interior serves as our cellar. With the hive there is neither cellar nor upper story: we are on one single level. Either way the query is identical. Given tumah in the dwelling, do the objects lying inside the well survive, and do those inside the hive survive, seeing that a kli immune to tumah lies across the top?
The verdict is no. Tamei. Nothing protrudes above the floor around the well, so no ohel partition exists here whatsoever, and a kli by itself has no rescuing power. A kli requires an ohel functioning at its side. The identical logic governs the hive. Granted, the hive has sides of its own, but a hive is no ohel; it is merely a kli. And yes, the basket covering it has sides, yet that also is nothing more than a kli. Some component of ohel has to join forces with the kli before tumah can be stopped.
A flat board is no kli, it is a roof
At this point comes a striking ruling: "im hayah neser chalak o seridah she'ein lah gapayim." A neser chalak is a smooth sawn plank set across the well's mouth. A seridah is likewise a wooden board, the sort used for kneading, and here it is a seridah "she'ein lah gapayim," carrying no raised edge along its borders. The mishnah's verdict: tahor.
The contrast is startling. Take a well inside a dwelling, or take a beehive standing there, and cap it with a colossal basket that cannot receive tumah: useless, the contents turn tamei. Cap that identical well or hive with an ordinary flat plank: tahor. The explanation is that the plank is no kli whatsoever. Its role is that of a roof, and a roof is matzil against tumah entirely by itself. It requires no ohel partition as a partner.
The mishnah then formulates the rule outright: "she'ein hakeilim matzilin im dafnos ha'ohel ad sheyehei lahen dofen." Should you be counting on a basket to preserve what lies in the well, it will fail unless some sort of partition is present. Should you be counting on a basket to preserve what lies in the hive, the outcome is the same. All because a basket is a kli. A flat plank, by contrast, is a covering, and whatever rests beneath a covering is untouched by tumah. Should you wish a kli to serve you as a roof, you must furnish an ohel to operate alongside it.
How much partition is required?
"Vechamah tehei hadofen?" What height must an ohel's partition reach to qualify as a partition, enabling the kli to combine with it? The answer: "tefach." Hardly a towering height, yet a complete tefach is demanded.
The Bartenura notes that a substantial chiddush is embedded here. That tefach of partition need not be part of the ohel. The pit's lip need not be what rises a tefach off the floor. Should the covering vessel's own side stand a tefach tall, that answers the requirement. (We alluded to this above so that it would not startle the reader now.)
The Tosfos Yom Tov articulates just how large this chiddush is. The mishnah's opening cases told us that the sides of keilim contribute nothing, and the reason is that they fail to reach the necessary height. It follows that everything stated until now, that the olive basket rescues neither the contents of the well nor the contents of the open hive, is limited to a basket whose sides fall short of a tefach. One complete tefach is what is needed, furnished by the ohel or else by the side of the kli laid over the opening.
Nor may that tefach be pieced together out of two fractions. "Hayah lo chatzi tefach mikan vechatzi tefach mikan, eino dofen." Suppose the pit's lip protrudes half a tefach above the floor and the covering basket adds another half a tefach of side: that combination is no partition. "Ad sheyehei lo tefach mimakom echad," the full tefach has to come from a single source, from the ohel or from the kli.
Arranging the cases
Let us set the possibilities in a row. An item lacking sides altogether, a bare plank, rescues whatever lies below it even when nothing at all protrudes above the floor of the ohel. A kli with sides a complete tefach tall likewise rescues, again with no lip on the pit. But a kli whose sides come to less than a tefach rescues the well's contents only when the well itself is ringed by a lip of at least a tefach standing above ground.
Two approaches to the olive basket
A difficulty arises. We described the chafishah as a huge basket loaded with olives, holding forty se'ah, which is precisely the reason it cannot receive tumah. If so, how could its side ever measure less than a tefach? Where would forty se'ah go?
The Tiferes Yisroel resolves it: the basket must have been exceedingly broad. Its sides are low, under a tefach, yet its width is such that forty se'ah of liquid finds room inside.
The Rambam takes another route: this basket does not hold forty se'ah at all. Then why is it immune to tumah? Because it is manufactured from substances that never receive tumah, such as gelalim, stone or earth. That is its exemption, and it may perfectly well stand lower than a tefach. And if it does possess a modest side, it will function as long as the pit has a tefach of lip emerging from the ground.
Once more, then. A covering with no sides whatsoever, a plain plank: fine, the well's contents remain tahor. Sides under a tefach: you require a tefach of partition supplied by the ohel. Sides of a complete tefach on the kli: fine, no ohel partition is called for, so a well in the ground capped by a basket whose sides measure a tefach keeps its contents safe. But a half-tefach contributed by the kli plus a half-tefach contributed by the pit's rim accomplishes nothing. The tefach must be whole, all of it in the ohel or all of it in the kli.