Eruvin 87a — the Talmud in English

He carves in it

a broad hole in its middle,

four

handbreadths

by four

handbreadths, and ten handbreadths remain around the hole on every side,

and he draws water

from the sea.

For we say: “Bend,” that is, we see those ten handbreadths that remain as though they were bent downward, like a suspended partition over the water; and we also say, “Draw down the partition,” for we see the suspended partitions as though they descend and close off down to the water. [See illustrations]

And the water opposite the hole becomes a private domain, and he may draw water from the sea and bring it into his house.

[But if the balcony was less than twenty-four handbreadths by twenty-four handbreadths, it is insufficient, since the hole must be four by four handbreadths, for otherwise it would be a partition made of less than four by four, which is not a partition. Likewise, ten handbreadths must remain around the hole on every side, as the measure of a partition.]

We see, then, that Rabbi Hanania ben Akavya was lenient with regard to abundant water, and did not require the partition to be recognizable as having been made for the sake of the water, for although the hole was made for the sake of the water, the partition was not made for the sake of the water [Gaon Yaakov; see also Ritba].

Abaye said to him: But perhaps it is not so

as you say: that Rabbi Yehuda and Rabbi Hanania ben Akavya expressed one position.

Rather, it must be said:

Rabbi Yehuda stated this only there

[in the Mishna],

but

because

he said: “Draw down the partition”

with regard to existing suspended partitions.

But “bend and draw down”—he did not. [See what Gaon Yaakov added in explanation of this.]

And Rabbi Hanania ben Akavya stated this only there,

to permit the residents of Tiberias to draw water from the Sea of Tiberias by means of this dispensation [as it will be said below, that it was for them that he permitted it],

but

because it deals

with the Sea of Tiberias,

which is not an entirely complete karmelit [a domain that has no partitions], since

it has banks

[a raised edge], which are like partitions surrounding it,

and towns and enclosures surround it

all around, and it is a domain that has partitions.

Nevertheless, it has the status of a karmelit by rabbinic enactment, since it is “an enclosure larger than a beit sataim that was not enclosed for residential purposes,” which the Sages made like a karmelit [but by Torah law it is an entirely complete private domain], as it is stated above, on page 23a.

And since that is so, they were lenient with regard to it, relying upon such a minor corrective measure.

But with regard to other bodies of water,

which are entirely complete karmeliyot,

he did not

make Rabbi Hanania lenient; and perhaps even in the case of the Mishna he would not be so lenient.

Abaye said: And according to Rabbi Hanania ben Akavya,

who holds that with regard to a balcony we say, “Bend and draw down”:

If the wall of the upper story [to which the balcony is attached] stood at the edge of the water, and

the

balcony was

adjacent to the wall, at a

distance

of less than three handbreadths

from the beginning of the balcony to the wall, the balcony need not be four cubits by four cubits.

Rather,

its length must be

of the balcony [along the upper story]

four cubits, and its width eleven

handbreadths

and something. [See illustration 1.] And this is the calculation: Since the upper story itself stands at the edge of the sea, the wall of the upper story is considered one partition of the balcony, and the balcony need only be sufficient for bending three partitions. We regard the width of the balcony, which is eleven handbreadths and something, as though the ten handbreadths distant from the upper story had been bent downward; thus we have a second partition parallel to the partition of the upper-story wall.

After the bending, one handbreadth and something of balcony width remains, along a length of four cubits, which are twenty-four handbreadths.

“Bend” the ten handbreadths on each side downward!

There is thus for you, on each side, a partition ten handbreadths high and one handbreadth and something wide, with the distance from the wall to the beginning of the partition less than three handbreadths.

And when you say “lavud,” and regard it as though the width of the partition extends to the wall, we have two additional partitions ten handbreadths high and four handbreadths wide.

What remains of the balcony after the “bends” is a length of four handbreadths by a width of one handbreadth and something.

Now we carve a hole across the entire remaining width of the balcony and along its entire remaining length; there is thus for you an empty space of four by four [with the gap between the balcony and the upper story].

And it turns out that an empty space of four by four is surrounded by partitions. [See illustration 2.]

And he lowers a bucket from the window of the upper story and draws water.

If it was upright,

that is, parallel to the wall of the upper story—which stands at the edge of the sea—he erected, at an exact distance of four handbreadths, a partition standing upright like the wall of the upper story,

then

its height must be

of that partition

ten handbreadths.

We then have two partitions—the partition of the upper story and the upright partition—ten handbreadths high, with a four-handbreadth hole between them.

We must complete two additional partitions connecting the two partitions,

and

therefore its

width

of the upright partition

must be six handbreadths and two somethings. [See illustration 1.]

And this is the calculation: Just as we say “bend” downward, here too we say “bend” that upright partition on both its sides toward the wall of the upper story!

And once one handbreadth and something has been bent on each side, there are between the bend and the wall of the upper story three handbreadths less something.

And when you say “lavud,” there is for you on each side of the hole a partition ten handbreadths high [the height of the upright partition] and four handbreadths wide.

And it stands adjacent to a hole whose length is four handbreadths [the width of the partition remaining after the “bending”], by four handbreadths [the distance between the wall and the parallel upright partition]. [See illustration 2.]

And he lowers a bucket from the window of the upper story and draws water.

If there was,

one of the walls of the upper story extending into the water from east to west for four handbreadths, and a second wall of the upper story running from north to south at the edge of the water, for a length of four handbreadths, and

the

balcony was

standing

and sunk

in

that

corner angle

of the water, and the two walls of the upper story were themselves the walls of the balcony.

It follows that only two partitions need be completed for the balcony. Therefore:

its height must be

of an upright partition—which he will erect parallel to the wall of the upper story running from north to south, at an exact distance of four handbreadths from this wall—

ten handbreadths, and its width two handbreadths and two somethings. [See illustration 3.]

And this is the calculation: Erect the partition from north to south, with its edge distant by three handbreadths less something from the wall running from east to west!

And there remains one handbreadth and something of it; together with the gap between it and the wall running from east to west, when you say “lavud,” this completes a third partition ten handbreadths high and four handbreadths wide.

There remains for you a width of one handbreadth and something of the upright partition, in order to “bend” it toward the wall running from north to south [a bend parallel to the wall running from east to west]. Thus you have a fourth partition, when you “add” to that bend, by means of “lavud,” those three handbreadths less something that separate the bend from the wall running from north to south.

And the bottom of the balcony is entirely hollow; thus you have a hole four by four, surrounded by four partitions ten handbreadths high. [See illustration 4.] And he lowers a bucket from a window in one of the walls of the upper story and draws water.

[Likewise, he may place that upright partition parallel to the wall running from east to west, and three handbreadths less something distant from the wall running from north to south.]

And now we ask: Since you said that when the balcony is adjacent to the wall of the upper story, the wall itself is like one partition, and it is also possible to join the empty space between the wall and the balcony, since we say “lavud,” this is difficult:

But if that is so, how can we understand the source that was taught: Rabbi Ḥananya ben Akavya says: In the case of a balcony that is four cubits by four cubits, one cuts into it an opening four by four

But if so, what about that which was taught: Rabbi Hanania ben Akavya says: A balcony that is four cubits by four cubits, he carves in it a hole four by four

handbreadths

and draws water—

how can you find it

that it needs this measure, which is sufficient to bend four partitions ten handbreadths high and to carve a hole four by four—and less than this is sufficient, according to what we have said?!

And we answer: We require four cubits only where

it is made

the balcony

like a mortar

[a mortar], that is, where the balcony is not adjacent to the wall but stands independently on four posts in the water. [See illustration.]

Therefore, it must have a measure sufficient to carve a hole four by four and to bend four partitions ten handbreadths high, which altogether are twenty-four handbreadths by twenty-four handbreadths, namely, four cubits by four cubits.

And after one has, as it were, bent the partitions, this balcony assumes the form of a mortar.

Mishna:

A water channel [a channel for the passage of water] when it is ten deep and four wide has the status of a karmelit, like the other seas and rivers.

And if the water channel passes through his courtyard—entering from one side of his courtyard to the other side of his courtyard through holes beneath the walls—it is prohibited to draw water from it into his courtyard, since he is drawing from a karmelit into a private domain.

But if there was a pool of water in his courtyard that does not leave the boundaries of the courtyard, it requires no correction whatsoever, even if it is very wide and deep, for everything within the partitions is included in the private domain [Mishna Berura].

A water channel that passes through a courtyard: one may not draw from it on Shabbat

[as was explained],

unless they made a partition for it high

at the entrance

ten handbreadths

within the water channel

at the entrance

of the water channel into the courtyard

and at the exit

from it. [See illustration.]

says: The wall of the courtyard that

Rabbi Yehuda—

according to his position in the earlier Mishna concerning a pit between two courtyards—

says: The wall

of the courtyard

that is

suspended

over it,

over the water channel,

is considered a partition!

Just as Rabbi Yehuda relies above on a suspended partition, although it was not made for the sake of the pit.

Rabbi Yehuda said,

to prove his position:

There was an incident involving a water channel

of water

of

the city

of Abel—

which had only the partitions of the courtyard suspended over it—

and they would draw water from it on Shabbat on the authority of the Elders!

The Sages said to him: There is no proof from there, for that is why they would draw water from it

because it did not have the measure

of a karmelit, namely, a depth of ten handbreadths and a width of four handbreadths.

but did not make one for it at the exit

Gemara:

The Sages taught: If they made for it

a partition for the water channel

at the entrance

to the courtyard,

but did not make one for it at the exit

from the courtyard,

or

they made one for it at the exit

from the courtyard,

but did not make one for it at the entrance,

since it is still connected to the water outside the courtyard, even what is within the courtyard becomes a karmelit.

Therefore:

one may not draw water from it on Shabbat, unless they made for it a partition of ten handbreadths at the exit and at the entrance,

so that, by means of the partitions, it appears as though the channel begins in this courtyard; therefore one may draw water from it on Shabbat.

Rabbi Yehuda says: The wall over it is considered a partition.

Rabbi Yehuda said: There was an incident involving a water channel that came from Abel to Sepphoris—

and it had only the partitions of the courtyard suspended over it—

and they would draw water from it on Shabbat on the authority of the Elders.

They said to him: Is there proof from there?!

That is why they drew water from it,

because it was not ten handbreadths deep,

or it did not have

a width of four

handbreadths.

Another beraita was taught: A water channel

that enters “an alleyway or a courtyard” [the wording of Rashi; see below in the name of Gaon Yaakov]

that passes between the windows,

that is, it passes between two rows of houses, and the windows of the houses open toward the channel in order to draw water from it, and they did not make partitions for it in order to permit it:

Less than three

handbreadths,

he lowers a bucket and draws water [the explanation proceeds]. [See illustration.]

Three

handbreadths,

he may not lower a bucket and draw water.

Rabban Shimon ben Gamliel says: Less than four

handbreadths,

he lowers a bucket and draws water.

Four

handbreadths,

he may not lower a bucket and draw water.

And now we explain the measure of three and four stated in the beraita:

With what are we dealing?

If you say

that the measure of three and four refers to

the water channel itself,

that is, the first tanna holds that if the width of the water channel is less than three handbreadths, it is not a karmelit, and therefore it is permitted to draw water from it even though no partitions were made for it; only from a width of three handbreadths does it become a karmelit, and it is prohibited. And Rabban Shimon ben Gamliel holds that its measure is four handbreadths.

That cannot be!

For it would be difficult:

But what of that which, when Rav Dimi came, he said in the name of Rabbi Yohanan: A karmelit is not less than

a width of

four

—

shall we say that he stated his teaching in accordance with a dispute among tannaim?!

Rather,

it is unavoidable that this measure of three and four does not refer to the width of the channel, which was as wide as the measure of a karmelit, but to the width of

its banks

[the walls of the water channel], which were three handbreadths high above the ground of the courtyard.

And

the subject of the beraita is

transferring [that is, lowering into the channel in two stages, from the window to the bank and from the bank to the channel].

And this is the explanation of the beraita: If the bank was “less than three” handbreadths wide, “he lowers a bucket” to the bank, and from the bank he lowers it again into the channel, and “draws water” from the channel. And he raises it from the channel to the bank, and from the bank to the window.

But if the bank was three handbreadths wide, “he may not lower a bucket and draw water,” even in two stages.

And according to Rabban Shimon ben Gamliel, it is permitted to transfer until the banks are four handbreadths wide; from four handbreadths onward, it is prohibited to transfer.

And the explanation of their dispute is as follows: A raised place three handbreadths high that stands between a private domain and a public domain, if it is four handbreadths wide and is not ten handbreadths high [so that it is not a private domain in its own right], is a karmelit.

And if it is not four handbreadths wide, it is not an independent domain [a karmelit], but is nullified to the two domains beside it, and it is permitted to remove from it both to the public domain and to the private domain; it is considered a “place exempt.”

The Rabbis hold that this applies when that place stands between a private domain and a public domain, which are significant domains by Torah law; therefore that place is nullified with respect to both of them, and its status as an independent domain is removed.

But if it stands between a private domain and a karmelit, which is only a domain by rabbinic law, the karmelit is not sufficiently significant for that place to be nullified even to it.

Therefore, it is nullified only to the significant private domain beside it, and it has the law of a private domain.

Only if that place is less than three handbreadths wide is it a place exempt.

Therefore, the banks of the channel that stand between the private domain [the courtyard] and the karmelit [the channel], if they are less than three handbreadths wide, are a place exempt; and when he removes from the window to the bank, it is merely like removing from a private domain to a place exempt, which is permitted.

And when he lowers it again from the bank to the channel, it is merely like removing from a place exempt to a karmelit, which is permitted.

But when they are three handbreadths wide, they have the law of a private domain, and it is prohibited to lower from the bank to the channel and to draw water.

And Rabban Shimon ben Gamliel holds: There is no distinction between a place adjacent to a domain by Torah law and one adjacent to a domain by rabbinic law. Rather, every place that does not contain four by four is a place exempt with regard to transferring from it to a karmelit, since this involves only a rabbinic prohibition.

And we raise an objection to this: Even though it is a place exempt, nevertheless it is prohibited.

For

when Rav Dimi came, he said in the name of Rabbi Yohanan: A place that does not contain four by four

—which is a place exempt, as we have said—

is permitted for the people of a private domain and for the people of a public domain to place burdens upon it

[to place their burdens upon it and to take them from it],

provided that they do not transfer between domains!

For it was not permitted to transfer from one domain to another, even if he places it in the middle, in a place exempt. Now this is difficult: How can you permit here transferring from a private domain to a karmelit by way of a place exempt?!

And we answer:

There,

in Rabbi Yohanan’s statement,

the domains are domains by Torah law,

for the place stands between a private domain and a public domain. It follows that when he transfers by way of a place exempt, “a Torah prohibition is performed by way of a place exempt” [Rashi]; therefore it is prohibited.