פירוש על עבודה זרה 99:18

The dispute can also not be where the stones are close because we would have to assume that they fell from the Mercurius.

They differ with regard to stones found within four cubits of the Mercurius, but ones we do not know came from the larger idol.
If there are three together then R. Ishmael holds that this is a “small Mercurius” which people make near a “big Mercurius” and therefore they are prohibited. Two is nothing and therefore it is permitted.

The rabbis do not perceive this to be a “small Mercurius.” So then all that matters is whether the stones are thought to have come from the Mercurius.

Above R. Yitzchak had said that if we know that the stones fell from the Mercurius, they are permitted. But neither the sages nor the rabbis in the mishnah say exactly that.
We should note that the terminology here is unusual. Usually the word ורמינהו means that two tannaitic sources disagree. But here it seems to be a difficulty from a tanna on and amora.

Rava here emends the first source so that it reads “found” and not “fell.”
There is still something strange here, for if the word is “found” then what is the sentence really telling us.

According to the mishnah, R. Ishmael seems to always allow two stones found next to a Mercurius. But according to a baraita, if they are within reach of the Mercurius, even two are prohibited.
Rava explains that if they are within one reach they are prohibited but if they are as far apart as two reaches they are permitted. This situation is defined as a case where there is a mound between the stones and the Mercurius.

Are stones to be considered a Mercurius if they are only right next to each other? A baraita states clearly that a Mercurius is three stones, one on top of two.
Rava explains that that baraita refers to the main part of a Mercurius, which indeed is one stone on two. But once the main part is built, additional stones can be added on the side.
I want to point out that the placing of stones on something is reminiscent of the Jewish practice of putting stones on a grave. Unnaturally placed stones are a way of people saying, “we were here.” Unfortunately, I was at a funeral today and so this sugya came to mind.

Yannai (Alexander Yannai) was a Hasmonean king who ruled between 103-76 B.C.E. According to this source, when his palace was destroyed, the Romans turned it into a palace for Mercurius. Then other idolaters came, conquered the place and did not worship Mercurius, so they used the stones to pave the ground. Thus we have stones that were once part of idols now being used to pave roads. Can a Jew walk there? Have they been annulled? Some rabbis did, and others did not.
As far as the historicity of this story goes, there may be some echo here of the fact that Israel was placed between the two great powers, the Seleucids in the north and the Ptolemies in the South. The rulers of Israel often found themselves caught between these two powers. I do not know if one of these powers worshipped Mercury while the other did not.

The rabbis who did not walk on these roads held that they were like idolatrous offerings that can never be annulled. Thus although they were taken down from the Temple, they still retain their idolatrous nature.

Yohanan refers to a very righteous person who would walk on those stones. This R. Menahem son of R. Simai was so holy that he would never look even at the image on a coin.

The one who did walk on these stones held that they are not to be considered “idolatrous offerings” because stones were not offered on the altar. Only things like actual idolatrous offerings cannot be annulled. Stones can be annulled.

The baraita distinguishes between an idolater who uses stones from a Mercurius to pave the roads, in which case a Jew can walk on the roads, and a Jew who uses these stones, in which case it is prohibited for the Jew to walk on the roads. But Rabbah b. Yirmiyah cannot explain this distinction. “Carpenter” here seems to be a nickname for a scholar, probably due to the high level of skill a carpenter needs to shape his material.