There are some that assert that baptism by immersion (submersion) into water was impractical for the three thousand described at Acts 2:14 because of doubts that water was plentiful enough to accomodate such a use.

Acts 2:41* Then they that gladly received his word were baptized: and the same day there were added unto them about three thousand souls.

A quick check into authoritative sources confirms that water in Jerusalem at that time (under Roman occupation no less, which Romans are famous for their aqueducts and their fondness for baths) was plentiful in the city, not to exclude any brooks or streams that might have on the outskirts at that time. See the references below.

From: THE WARS OF THE JEWS, BOOK 5, CHAPTER 4, THE DESCRIPTION OF JERUSALEM, p. 1650.
"There were, moreover, several groves of trees, and long walks through them, with deep canals, and cisterns, that in several parts were filled with brazen statues, through which the water ran out."

From: INTERNATIONAL STANDARD BIBLE ENCYCLOPEDIA, VOL. 6, J-KUSHAIAH, Jerusalem, VII. ANTIQUARIAN REMAINS CONNECTED WITH THE WATER-SUPPLY. pgs. 211-219.

VII. ANTIQUARIAN REMAINS CONNECTED WITH THE WATER-SUPPLY.

In a city like Jerusalem, where the problem of a water-supply must always have been one of the greatest, it is only natural that some of the most ancient and important works should have centered round it. The three sources of supply have been

(1) springs,
(2) cisterns,
(3) aqueducts.

1. Gihon: The Natural Spring:

(1) The natural springs have been described in II, 3; but connected with them, and especially with the city’s greatest and most venerated source, the Gihon, there are certain antiquarian remains of great interest.

(a) The "Virgin’s Fount," ancient Gihon, arises, as has been described (II, 3), in a rocky cleft in the Kidron valley bottom; under natural conditions the water would run along the valley bed, now deeply buried under debris of the ancient city, and doubtless when the earliest settlers made their dwellings in the caves (which have been excavated) on the sides of the valley near the spring, they and their flocks lived on the banks of a stream of running water in a sequestered valley among waterless hills. From, however, a comparatively early period – at the least 2000 BC – efforts were made to retain some of the water, and a solid stone dam was built which converted the sources into a pool of considerable depth. Either then, or somewhat later, excavations were made in the cliffs overhanging the pool, whereby some at least of these waters were conducted, by means of a tunnel, into the heart of the southeastern hill, "Ophel," so that the source could be reached from within the city walls. There are today two systems of tunnels which are usually classed as one under the name of the "Siloam aqueduct," but the two systems are probably many centuries apart in age.

2. The Aqueduct of the Canaanites:

The older tunnel begins in a cave near the source and then runs westward for a distance of 67 ft.; at the inner end of the tunnel there is a perpendicular shaft which ascends for over 40 ft. and opens into a lofty rock-cut passage which runs, with a slight lateral curvature, to the North, in the direction of the surface. The upper end has been partially destroyed, and the roof, which had fallen in, was long ago partially restored by a masonry arch. At this part of the passage the floor is abruptly interrupted across its whole width by a deep chasm which Warren partially excavated, but which Parker has since conclusively shown to end blindly. It is clear that this great gallery, which is 8 to 9 ft. wide, and in places as high or higher, was constructed (a natural cavern possibly utilized in the process) to enable the inhabitants of the walled-in city above it to reach the spring. It is in fact a similar work to the great water-passage at GEZER (which see), which commenced in a rock-cut pit 26 ft. deep and descended with steps, to a depth of 94 ft. 6 inches below the level of the rock surface; the sloping passage was 23 ft. high and 13 ft. broad. This passage which could be dated with certainty as before 1500 BC, and almost certainly as early as 2000 BC, was cut out with flint knives and apparently was made entirely to reach a great underground source of water.

3. Warren’s Shaft:

The discovery of this Gezer well-passage has thrown a flood of light upon the "Warren’s Shaft" in Jerusalem, which would appear to have been made for an exactly similar purpose. The chasm mentioned before may have been an effort to reach the source from a higher point, or it may have been made, or later adapted, to prevent ingress by means of the system of tunnels into the city. This passage is in all probability the "watercourse" (tsinnor) of 2 Sam 5:8 up which, apparently, Joab and his men (1 Ch 11:6) secretly made their way; they must have waded through the water at the source, ascended the perpendicular shaft (a feat performed in 1910 by some British officers without any assistance from ladders), and then made their way into the heart of the city along the great tunnel. Judging by the similar Gezer water tunnel, this great work may not only have existed in David’s time, but may have been constructed as much as 1,000 years before.

4. Hezekiah’s "Siloam" Aqueduct:

The true Siloam tunnel is a considerably later work. It branches off from the older aqueduct at a point 67 ft. from the entrance, and after running an exceedingly winding course of 1,682 ft., it empties itself into the Pool of Siloam (total length 1,749 ft.). The whole canal is rock cut; it is 2 to 3 ft. wide, and varies in height from 16 ft. at the south end to 4 ft. 6 inches at the lowest point, near the middle. The condition of this tunnel has recently been greatly changed through Captain Parker’s party having cleared out the accumulated silt of centuries; before this, parts of the channel could be traversed only with the greatest difficulty and discomfort. The primitive nature of this construction is shown by the many false passages made, and also by the extensive curves which greatly add to its length. This latter may also be partly due to the workmen following lines of soft strata. M. Clermont-Ganneau and others have thought that one or more of the great curves may have been made deliberately to avoid the tombs of the kings of Judah. The method of construction of the tunnel is narrated in the Siloam Inscription (see SILOAM). It was begun simultaneously from each end, and the two parties met in the middle. It is a remarkable thing that there is a difference of level of only one foot at each end; but the lofty height of the southern end is probably due to a lowering of the floor here after the junction was effected. It is practically certain that this great work is that referred to in 2 Ki 20:20: "Now the rest of the acts of Hezekiah, and all his might, and how he made the pool, and the conduit, and brought water into the city, are they not written in the book of the chronicles of the kings of Judah?" And in 2 Ch 32:30: "This same Hezekiah also stopped the upper spring of the waters of Gihon, and brought them straight down on the west side of the city of David."

5. Other Aqueducts at Gihon:

In addition to these two conduits, which have a direct Scriptural interest, there are remains of at least two other aqueducts which take their origin at the Virgin’s Fount – one a channel deeply cut in rock along the western sides of the Kidron valley, found by Captain Parker, and the other a built channel, lined with very good cement, which takes its rise at a lower level than any of the other conduits close to the before-mentioned rocky cleft from which the water rises, and runs in a very winding direction along the western side of the Kidron. This the present writer has described in PEFS, 1902. One of these, perhaps more probably the former, may be the conduit which is referred to as Shiloah ([shiloach]), or "conducted" (Isa 8:6), before the construction of Hezekiah s work (see SILOAM).

There are other caves and rock-cut channels around the ancient Gihon which cannot fully be described here, but which abundantly confirm the sanctity of the site.

6. Bir Eyyub:

(b) Bir Eyyub has a depth of 125 ft.; the water collects at the bottom in a large rock-hewn chamber, and it is clear that it has been deepened at some period, because at the depth of 113 ft. there is a collecting chamber which is now replaced by the deeper one. Various rock-cut passages or staircases were found by Warren in the neighborhood of this well.

7. Varieties of Cisterns:

(2) The cisterns and tanks. – Every ancient site in the hill country of Palestine is riddled with cisterns for the storage of rain water. In Jerusalem for very many centuries the private resident has depended largely upon the water collected from the roof of his house for all domestic purposes. Such cisterns lie either under or alongside the dwelling. Many of the earliest of these excavations are bottle-shaped, with a comparatively narrow mouth cut through the hard Mizzeh and a large rounded excavation made in the underlying Melekeh (see II, 1 above). Other ancient cisterns are cavities hewn in the rock, of irregular shape, with a roof of harder rock and often several openings. The later forms are vaulted over, and are either cut in the rock or sometimes partially built in the superlying rubbish. For more public purposes large cisterns were made in the Charam, or temple-area. Some 3 dozen are known and planned; the largest is calculated to contain 3,000,000 gallons. Such structures were made largely for the religious ritual, but, as we shall see, they have been supplied by other sources than the rainfall. In many parts of the city open tanks have been constructed, such a tank being known in Arabic as a birkeh, or, followed by a vowel, birket. With most of these there is considerable doubt as to their date of construction, but probably none of them, in their present form at any rate, antedates the Roman period.

8. Birket Israel:

Within the city walls the largest reservoir is the Birket Israel which extends from the northeastern angle of the Charam westward for 360 ft. It is 125 ft. wide and was originally 80 ft. deep, but has in recent years been largely filled up by the city’s refuse. The eastern and western ends of this pool are partially rock-cut and partly masonry, the masonry of the former being a great dam 45 ft. thick, the lower part of which is continuous with the ancient eastern wall of the temple-area. The sides of the pool are entirely masonry because this reservoir is built across the width of the valley referred to before (III, 2) as "St. Anne’s Valley." Other parts of this valley are filled with debris to the depth of 100 ft. The original bottom of the reservoir is covered with a layer of about 19 inches of very hard concrete and cement. There was a great conduit at the eastern end of the pool built of massive stones, and connected with the pool by a perforated stone with three round holes 5 1/2 inches in diameter. The position of this outlet shows that all water over a depth of 22 ft. must have flowed away. Some authorities consider this pool to have been pre-exilic. By early Christian pilgrims it was identified as the "Sheep Pool" of Jn 5:2, and at a later period, until quite recent times, it was supposed to have been the Pool of Bethesda.

9. Pool of Bethesda:

The discovery, a few years ago, of the long-lost Piscina in the neighborhood of the "Church of Anne," which was without doubt the Pool of Bethesda of the 5th century AD, has caused this identification to be abandoned.

See BETHESDA.

10. The Twin Pools:

To the West of the Birket Israel are the "twin pools" which extend under the roadway in the neighborhood of the "Ecce Homo" arch. The western one is 165 ft. by 20 ft. and the eastern 127 ft. by 20 ft. M. Clermont-Ganneau considers them to be identical with the Pool Struthius of Josephus (BJ, V, xi, 4), but others, considering that they are actually made in the fosse of the Antonia, give them a later date of origin. In connection with these pools a great aqueduct was discovered in 1871, 2 1/2-3 ft. wide and in places 12 ft. high, running from the neighborhood of the Damascus Gate – but destroyed farther north – and from the pools another aqueduct runs in the direction of the Charam.

11. Birket Hammam el Batrak:

On the northwestern hill, between the Jaffa Gate and the Church of the Sepulchre there is a large open reservoir, known to the modern inhabitants of the city as Birket Kammam el Batrak, "the Pool of the Patriarch’s Bath." It is 240 ft. long (North to South), 144 ft. broad and 19-24 ft. deep. The cement lining of the bottom is cracked and practically useless. The eastern wall of this pool is particularly massive, and forms the base of the remarkably level street Karet en Nasara, or "Christian Street"; it is a not improbable theory that this is actually a fragment of the long-sought "second" wall. If so, the pool, which is proved to have once extended 60 ft. farther north, may have been constructed originally as part of the fosse. On the other hand, this pool appears to have been the Amygdalon Pool, or "Pool of the Tower" (berekhath ha-mighdalin), mentioned by Josephus (Jewish Wars, V, xi, 4), which was the scene of the activities of the 10^th legion, and this seems inconsistent with the previous theory, as the events described seem to imply that the second wall ran outside the pool. The popular travelers’ name, "Pool of Hezekiah," given to this reservoir is due to theory, now quite discredited, that this is the pool referred to in 2 Ki 20:20, "He made the pool, and the conduit, and brought water into the city." Other earlier topographists have identified it as the "upper pool" of Isa 7:3; 36:2.

12. Birket Mamilla:

The Birket Kammam el Batrak is supplied with water from the Birket Mamilla, about 1/2 mile to the West. This large pool, 293 ft. long by 193 ft. broad and 19 1/2 ft. deep, lies in the midst of a large Moslem cemetery at the head of the Wady Mes, the first beginning of the Wady er Rababi (Hinnom). The aqueduct which connects the two pools springs from the eastern end of the Birket Mamilla, runs a somewhat winding course and enters the city near the Jaffa Gate. The aqueduct is in bad repair, and the water it carries, chiefly during heavy rain, is filthy. In the Middle Ages it was supposed that this was the "Upper Pool of Gihon" (see GIHON), but this and likewise the "highway of the FULLER’S FIELD" (which see) are now located elsewhere. Wilson and others have suggested that it is the "Serpent’s Pool" of Josephus (Jewish Wars, V, iii, 2). Titus leveled "all the places from Scopus to Herod’s monument which adjoins the pool called that of the Serpent." Like many such identifications, there is not very much to be said for or against it; it is probable that the pool existed at the time of the siege. It is likely that this is the Beth Memel of the Talmud (the Babylonian Talmud, `Erubin 51 b; Sanhedrin 24 a; Bere’shith Rabba’ 51).

13. Birket es Sultan:

The Birket es Sultan is a large pool – or, more strictly speaking, enclosure – 555 ft. North and South by 220 ft. East and West. It is bounded on the West and North by a great curve of the low-level aqueduct as it passes along and then across the Wady er Rababi. The southern side consists of a massive dam across the valley over which the Bethlehem carriage road runs. The name may signify either the "great" pool or be connected with the fact that it was reconstructed in the 16th century by the sultan Suleiman ibn Selim, as is recorded on an inscription upon a wayside fountain upon the southern wall. This pool is registered in the cartulary of the Holy Sepulchre as the Lacus Germani, after the name of a knight of Germanus, who built or renovated the pool in 1176 AD. Probably a great part of the pool is a catchment area, and the true reservoir is the rock-cut birkeh at the southern end, which has recently been cleaned out. It is extremely difficult to believe that under any conditions any large proportion of the whole area could ever have even been filled. Today the reservoir at the lower end holds, after the rainy season, some 10 or 12 ft. of very dirty water, chiefly the street drainage of the Jaffa road, while the upper two-thirds of the enclosure is used as a cattle market on Fridays. The water is now used for sprinkling the dusty roads in dry seasons.

The Pool of Siloam and the now dry Birket el Kamra are described under SILOAM (which see).

There are other tanks of considerable size in and around the city, e.g. the Birket Sitti Miriam, near "St. Stephen’s Gate," an uncemented pool in the Wady Joz, connected with which there is a rockcut aqueduct and others, but they are not of sufficient historical importance to merit description here.

14. "Solomon’s Pools":

(3) The conduits bringing water to the city from a distance are called the "high-level" and "low-level" aqueducts respectively, because they reached the city at different levels – the former probably somewhere near the present Jaffa Gate, the latter at the temple-platform.

15. Low-Level Aqueduct:

The low-level aqueduct which, though out of repair, can still be followed along its whole course, conveyed water from three great pools in the Wady `Artas, 7 miles South of Jerusalem. They are usually called "Solomon’s pools," in reference perhaps partly to Eccl 2:6: "I made me pools o water, to water therefrom the forest where trees were reared," but as any mighty work in Palestine is apt to be referred to the wise king of Israel, much stress cannot be laid on the name. These three storage reservoirs are constructed across the breadth of the valley, the lowest and largest being 582 ft. long by 177 ft. broad and, at the lowest end, 50 ft. deep. Although the overflow waters of `Ain es Saleh, commonly known as the "sealed fountain" (compare Song 4:12), reach the pools, the chief function was probably to collect the flood waters from the winter rains, and the water was passed from tank to tank after purification. There are in all four springs in this valley which supply the aqueduct which still conveys water to Bethlehem, where it passes through the hill by means of a tunnel and then, after running, winding along the sides of the hill, it enters another tunnel now converted into a storage tank for Jerusalem; from this it runs along the mountain sides and along the southern slopes of the site of Jerusalem to the Charam. The total length of this aqueduct is nearly 12 miles, but at a later date the supply was increased by the construction of a long extension of the conduit for a further 28 miles to Wady `Arrub on the road to Hebron, another 5 miles directly South of the pools. Here, too, there is a reservoir, the Birket el `Arrub, for the collection of the floodwater, and also several small springs, which are conducted in a number of underground rock-cut channels to the aqueduct. The total length of the low-level aqueduct is about 40 miles, and the fall in level from Birket el `Arrub (2,645 ft. above sea-level) at its far end to el Kas, the termination in the Charam Jerusalem (2,410 ft. above sea-level), is 235 ft.

16. High-Level Aqueduct:

The high-level aqueduct commences in a remarkable chain of wells connected with a tunnel, about 4 miles long, in the Wady Biar, "the Valley of Wells." Upward of 50 wells along the valley bottom supplied each its quotient; the water thence passed through a pool where the solid matter settled, and traversed a tunnel 1,700 ft. long into the `Artas valley. Here, where its level was 150 ft. above that of the low-level aqueduct, the conduit received the waters of the "sealed fountain," and finally "delivered them in Jerusalem at a level of about 20 ft. above that of the Jaffa Gate" (Wilson). The most remarkable feature of this conduit is the inverted siphon of perforated limestone blocks, forming a stone tube 15 inches in diameter, which carried the water across the valley near Rachel’s Tomb.

17. Dates of Construction of These Aqueducts:

On a number of these blocks, Latin inscriptions with the names of centurions of the time of Severus (195 AD) have been found, and this has led many to fix a date to this great work. So good an authority as Wilson, however, considers that these inscriptions may refer to repairs, and that the work is more probably Herodian. Unless the accounts of Josephus (Jewish Wars, V, iv, 4; II, xvii, 9) are exaggerated, Herod must have had some means of bringing abundant running water into the city at the level obtained by this conduit. The late Dr. Schick even suggested a date as early as Hyrcanus (135-125 BC). With regard to the low-level aqueduct, we have two definite data. First Josephus (Ant., XVIII, iii, 2) states that Pontius Pilate "undertook to bring a current of water to Jerusalem, and did it with the sacred money, and derived the origin of the stream from the distance of 200 furlongs," over 22 miles; in Jewish Wars, II, ix, 4 he is said to have brought the water "from 400 furlongs" – probably a copyist’s error. But these references must either be to restorations or to the extension from Wady `Arrub to Wady `Artas (28 miles), for the low-level aqueduct from the pools to Jerusalem is certainly the same construction as the aqueduct from these pools to the "Frank Mountain," the Herodium, and that, according to the definite statements of Josephus (Ant., XV, ix, 4; BJ, I, xxi, 10), was made by Herod the Great. On the whole the usual view is that the high-level aqueduct was the work of Severus, the low-level that of Herod, with an extension southward by Pontius Pilate.

Jerus still benefits somewhat from the low-level aqueduct which is in repair as far as Bethlehem, though all that reaches the city comes only through a solitary 4-inch pipe. The high-level aqueduct is hopelessly destroyed and can be traced only in places; the wells of Wady Biar are choked and useless, and the long winding aqueduct to Wady `Arrub is quite broken.