Taharos, Chapter 8, Mishnah 8. As the chapter draws toward its close, the Mishnah leaves behind the specific scenarios it had been discussing and turns to broader rules of tumah. Our mishnah illustrates that shift nicely: the question is when several small amounts of tamei food join together to reach the required measure, and the answer hangs on something remarkably mundane, namely whether the kneading trough is resting flat or propped at an angle.
The case
"Arivah shehi ketafres": an arivah is the trough in which dough is worked, and here it is positioned on a slant. "Vehatzek milma'alan": the dough sits at the raised end. "Umashkeh tofe'ach milmatan": beneath it lies wet liquid, trickling downward along the slope.
"Shalosh chatichos kabeitzah, einan mitztarfos." Picture three separate lumps of tamei dough whose combined volume comes to a kabeitzah, the bulk of an egg, the smallest amount by which food conveys tumah. Even so, the Mishnah rules that they do not join. The reason lies in the tilt: liquid sliding down an incline is not treated as one continuous substance uniting the lumps into a single kabeitzah. It merely happens to be underneath three independent pieces, none of which reaches a kabeitzah by itself; and if you add in the bit of liquid caught between any two of them, those two still do not amount to an egg's bulk. Consequently the dough does not convey tumah to the liquid below, and since the liquid stays clean, it never passes tumah on to the trough.
Two pieces
"Ushtayim mitztarfos": with only two lumps the ruling reverses, they do combine, and they make the liquid beneath them tamei. The liquid resting in the space between them touches each of the two, so the pair is linked and together amounts to a kabeitzah of tamei food.
Rabbi Yose holds otherwise: "af shtayim einan mitztarfos, ela im kein hayu rotztzos mashkeh." In his view even a pair fails to combine unless they squeeze the liquid, that is, unless they are set closely enough together to trap the liquid between them and prevent it from sliding away toward the bottom of the vessel. Only liquid held in place that way genuinely connects. The halachah is not decided according to Rabbi Yose.
When the liquid is standing
"Ve'im hayah mashkeh omeid": should the liquid be at rest instead of flowing, the trough lying level so that the liquid spreads out underneath the entire mass of dough, then, says the Mishnah, "afilu ke'ein chardal mitztaref": even fragments as tiny as a mustard seed, far smaller than the lumps discussed until now, join with one another. Motionless liquid is in contact with each of those specks and binds them, so that collectively they reach an egg's bulk of tamei food. That suffices to make the liquid tamei, and the tamei liquid then makes the kneading trough tamei as well.
Rabbi Dosa ends the mishnah with an objection: "ochel parud eino mitztaref." Food that has broken into separate crumbs can never be added together to form a measure, whatever the liquid may be doing. Here as well, we do not rule like Rabbi Dosa.
Everything, then, depends on how the trough is set. Raise one end and the liquid becomes a runoff that binds nothing at all; set it flat and that very same liquid turns into a bond firm enough to draw even mustard-seed specks together into a complete kabeitzah of tumah.