However, if a plumb line drops straight down, i.e., if the canyon wall is very steep, he measures the width of the canyon properly at the bottom of the canyon, without taking its walls into account.

The Gemara asks: And what is the depth of a canyon that may be spanned if it is not more than fifty cubits wide? Rav Yosef said: Up to two thousand cubits; but if it is deeper than that, the slope must be measured as well.

Abaye raised an objection from the following baraita: If a canyon is up to one hundred cubits deep and up to fifty cubits wide, one may span it; and if not, one may not span it. How could Rav Yosef say that the canyon may be spanned if its depth is less than two thousand cubits? The Gemara answers: He stated his opinion in accordance with the opinion of Aḥerim; as it was taught in a baraita: Aḥerim say: Even if the canyon is two thousand cubits deep and fifty cubits wide, one may span it.

The Gemara cites an alternate version of the previous discussion. Some say that Rav Yosef said: Even if the canyon is more than two thousand cubits deep, it may be spanned. The Gemara asks: In accordance with whose opinion did Rav Yosef say this? It is not in accordance with the opinion of the first tanna, and it is not in accordance with the opinion of the Aḥerim.

The Gemara answers: There, where the tanna’im disagree about the depth of a canyon that may be spanned, they refer to a case where a plumb line does not drop straight down and therefore there is reason to measure the slope. Here, however, where Rav Yosef says that the canyon may be spanned even if it is more than two thousand cubits deep, he is referring to a case where a plumb line drops straight down.

The Gemara asks: And where a plumb line does not drop straight down, how much must it extend from the top of the canyon in order for the wall of the canyon to be considered a slope rather than a vertical wall? Avimi said: Four cubits. If the bed of the canyon lies four cubits beyond the top edge of the canyon, the wall is sloped and must be included in the measurement. And similarly, Rami bar Ezekiel taught, based upon a baraita, that the maximum run is four cubits.

We learned in the mishna: If he reached a hill, he does not measure its height, but rather he spans the hill as if it were not there and then resumes his measurement. Rava said: They taught this halakha only with regard to a hill that has an incline of ten handbreadths within a run of four cubits. However, with regard to a gentler hill, e.g., one that has an incline of ten handbreadths within five cubits, one must measure the hill properly, i.e., he must include the slope itself in his measurement.

The Gemara notes that Rav Huna, son of Rav Natan, teaches a lenient formulation of this halakha: Rava said that they only taught this halakha with regard to a hill that has an incline of ten handbreadths within a run of five cubits. However, with regard to a steeper hill that has an incline of ten handbreadths within four cubits, one need not take any precise measurements; instead, he estimates the length of the hill, and then leaves and continues measuring from the other side.

We learned in the mishna that one may measure a canyon or hill located within the Shabbat limit, provided that one does not go out beyond the limit. The Gemara asks: What is the reason for this restriction? Rav Kahana said: It is a decree, lest people say: The measurement of the Shabbat limit comes to here. Since people know that he set out to measure the Shabbat limit, if they see him measuring in a certain spot they will assume that the area is included in the Shabbat limit.

We learned in the mishna: If, due to the width of the canyon or hill, one cannot span it, he may pierce it. The Sages taught a baraita which explains this procedure: How does one figuratively pierce a hill? Two people hold the two ends of a measuring rope. The one who is lower down on the hill holds the rope at the level of his heart while the one who is higher holds it at the level of his feet, and they proceed to measure in this fashion. Abaye said: Based on tradition, we hold that one may pierce only with a rope of four cubits.

Rav Naḥman said that Rabba bar Avuh said: Based on tradition, we hold that one may not pierce when measuring distances for the rite of the beheaded heifer. This rite is practiced when a murder victim is found, and it is not known who killed him. Judges measure the distance from the location of the corpse to the nearest town, in order to determine which town must perform the rite (Deuteronomy 21). Similarly, one may not pierce when measuring distances with regard to cities of refuge, in order to determine the boundaries within which an accidental murderer is protected from the blood redeemer (Numbers 32). Because these measurements are from the Torah, indirect methods of measurement are insufficient. The area must be measured as though it were flat.

MISHNA: One may measure the Shabbat limit only with an expert surveyor. If it is discovered that the surveyor extended the limit in one place and reduced it in another place, so that the line marking the Shabbat limit is not straight, one accepts the measurement of the place where he extended the limit and straightens the limit accordingly. Similarly, if the surveyor extended the limit for one and reduced it for another, one accepts the extended measurement.

And furthermore, even a gentile slave and even a gentile maidservant, whose testimonies are generally considered unreliable, are trustworthy to say: The Shabbat limit extended until here; as the Sages did not state the matter, the laws of Shabbat limits, to be stringent, but rather to be lenient. The prohibition to walk more than two thousand cubits is rabbinic in origin and is therefore interpreted leniently.