Negaim, Chapter 4, Mishnah 1. The Mishnah has been mapping out the periods of hesger for each kind of nega: the four kinds that settle on a human body (the or basar, the shechin and michvah, the netakim of the hairy areas of the rosh and zakan, and korachas and gabachas on the bald scalp), and then the tzara'as of begadim, garments. Now the Mishnah reaches the last of the categories, the tzara'as of batim, houses, and closes with a rule that fixes the outer limits of the whole system.
The Mishnah states: "HaBatim mitam'im bishloshah shavu'os uvishloshah simanim: b'yerakrak, uva'adamdam, uvifisyon." Houses contract tumah over three weeks and through three signs: the deep green, the deep red, and spreading. Recall how brief some of the other periods are; the shechin and the michvah are given only a single week. Houses stand at the opposite end of the scale, with three.
The Mishnah then spells out precisely when each siman takes effect. For the two colors: "B'yerakrak uva'adamdam bit'chilah", at the very first examination, and then "b'sof shavua rishon", once the first week has run its course, "uv'sof shavua sheni", once the second has run its course, "uv'sof shavua shlishi", once the third has run its course, "ule'achar haptur", and even after the house has already been released. Spreading appears on the identical list, minus the opening item: "Uv'fisyon", and through pisyon, at the conclusion of each of the three weeks and after the release as well.
How the simanim operate
The two colors, yerakrak and adamdam, are effective bit'chilah, from the Kohen's very first look. As with begadim, this does not mean that the house becomes muchlat on the spot. It means that such a discoloration is enough to set the process in motion: seeing deep green or deep red on his opening visit, the Kohen shuts the house for seven days. (And here it is worth remembering that the Kohen travels out to the nega, since a house plainly cannot be brought to him.) Pisyon behaves differently. It carries no weight at that opening examination; should the Kohen walk in and find the discoloration already spreading, he does not declare tumah on that basis. Spreading counts only once a week of hesger has been completed, whether the first, the second, or the third (assuming no spreading had shown itself earlier), and it counts again after the house was dismissed as tahor. The colors, too, remain in force after the p'tur: if the Kohen has already sent the owner off with a clean house and the yerakrak or adamdam reappears, the house is muchlat.
The Mishnah counts the days: "Umitamei bishloshah shavu'os, shehein tish'ah asar yom": three weeks, which come to nineteen days, since each new week begins on the seventh day of the previous one.
And then the summary rule: "Ein b'nega'im pachos mishavua echad, v'lo yoseir al sheloshah shavu'os." No nega is examined for less than one week, which is the case of the shechin and the michvah, and none for more than three weeks, which is the case of houses.
The Rambam on the three weeks of a house
Because nigei batim are intricate, it is worth reading how the Rambam frames them in his Peirush HaMishnayos, without yet entering into every detail of what happens week by week; that belongs to the later perek that treats nigei batim directly.
He opens by describing the appearance of a house nega: depressed streaks whose color is either deep green or deep red. Of such a nega the Torah writes that it "pasa", spread, within the walls of the dwelling. The Rambam then makes a point that is easy to overlook: nowhere do the pesukim state these three weeks openly. They are arrived at by joining what the verses do say to the explanation handed down by tradition. That, one might add, is exactly the reason for holding back the full week-by-week picture until its proper setting.
The Rambam traces it out. The Torah says "v'hisgir es habayis shiv'as yamim": the house is shut up and watched for seven days. Then the Torah commands that if the Kohen finds after that week that the discoloration has spread, "yachlotz osan ha'avanim", the stones are pried out, the plaster around them is scraped away, and the house is plastered a second time, as the pasuk says "v'afar acher yikach v'tach es habayis." Then the Torah continues: "uva hakohein v'ra'ah v'hinei pasa hanega babayis."
Here the received tradition enters, as recorded in the Sifra. That verse, the Rambam explains, is not describing a nega that came back to the house after the fresh plastering was applied. In that situation no spreading whatsoever is required: the moment tzara'as shows itself in the house again, the house comes down. The source is a gezeirah shavah resting on the expression tzara'as mam'eres, a malignant, eating tzara'as. Since the tzara'as mam'eres written concerning begadim renders a returning nega tamei with no spreading needed, the tzara'as mam'eres written concerning batim accomplishes the same.
What, then, is the verse about the nega having spread describing? The stage that follows the third week. Lay the order out: tzara'as was found in the house, it was closed up for seven days, and nothing spread; it was closed for a second seven days, and this time the nega neither faded nor stayed put but grew. Now the stones are pulled out, replacements are set into the wall, fresh plaster is applied, and a third hesger of seven days begins. Should the nega spread after that, the building is torn down, "horso oso", and it is this final stage that the pasuk is addressing.
The Rambam signs off with an undertaking: a fuller treatment of nigei batim will come where the subject belongs, later on in this masechta. We will do the same, saving the mechanics for the perek devoted to houses. What this Mishnah has handed us is the framework: three simanim, three weeks, nineteen days, and the outer borders of the entire system of nega'im, no examination shorter than a single week and none longer than three.