FI´STULA
FI´STULA (σωλήν), a waterpipe. Vitruvius (8.6) distinguishes three modes of conveying water: by channels of masonry (per canales
Fistulae. (From Middleton's Rome. ）

structiles), by leaden pipes (fistulis plumbeis), and by earthen pipes (tubulis fictilibus).^{1} Of these two sorts of pipes the leaden were the more commonly used in distributing the water from aqueducts: though drainingpipes of earthenware were extensively used, as in modern times, and many examples have been found. The lead pipes were made by rolling thick plates, in lengths of ten feet, round a wooden core; the edges were then brought together and soldered with melted lead. The resulting form was not perfectly cylindrical, but pearshaped in section, having a sort of ridge where the edges of the plate were somewhat roughly joined together. (See fig. D.) The pipes were made of much thicker lead than is the custom now, as much as 20 lbs. to the square foot in existing specimens (Middleton, p. 460); at points where there was heavy hydraulic pressure they were strengthened or replaced by bronze pipes, and the taps and stopcocks (epistomia, figs. E, F) were likewise of that metal.
In the manufacture of these pipes, particular attention was paid to the bore and to the thickness. The accounts of Vitruvius, Frontinus, and other writers are not in perfect accordance; but it appears, from a comparison of them, that two different systems of measurement were adopted,namely, either by the width of the plate of lead (lamina or lamna) before it was bent into the shape of a pipe, or by the internal diameter or bore (lumen) of the pipe when formed. The former is the system adopted by Vitruvius (l.c. § 4); according to him, the leaden plates were cast of a length not less than ten feet, and of a width containing an exact number of digits (sixteenths of a foot), which number was of course different for different sized pipes; and then the sizes of the pipes were named from the number of digits in the width of the plates, as in the following table, where the numbers on the right hand indicate the number of pounds which Vitruvius assigns to each tenfeet length of pipe :
Centenaria, from a plate  100 digits wide:  1200 lbs. 
Octogenaria from a plate  80 digits wide:  960 lbs. 
Quinquagenaria from a plate  50 digits wide:  600 lbs. 
Quadragenaria from a plate  40 digits wide:  480 lbs. 
Tricenaria from a plate  30 digits wide:  360 lbs. 
Vicenaria from a plate  20 digits wide:  240 lbs. 
Quindena from a plate  15 digits wide:  180 lbs. 
Dena from a plate  10 digits wide:  120 lbs. 
Octona from a plate  8 digits wide:  96^{2} lbs. 
Quinaria from a plate  5 digits wide:  60 lbs. 
If this scale is correct, the thickness of the plates must have been the same for pipes of all sizes; namely, such that each strip of lead, ten feet long and one digit wide, weighed twelve pounds. The account of Vitruvius is followed by Pliny (Plin. Nat. 31. § § 57, 58) and Palladius (9.12).
Frontinus, who enters into the subject much more minutely, objects to the system of Vitruvius as too indefinite, on account of the variation which is made in the shape of the pipe in bending up the plate of lead; and he thinks it more probable that the names were derived from the size of the internal diameters, reckoned [p. 1.863]in quadrantes (the unit being the digit), that is, in quarters of a digit; so that the Quinaria had a diameter of fivefourths of a digit, or 1 1/4 digit, and so on, up to the Vicenaria, above which the notation was altered, and the names were no longer taken from the number of linear quarters of a digit in the diameter of the pipe, but from the number of square quarters of a digit in its area. The usual limits of size were from the Quinaria, adopted by Frontinus as the standard measure (modulus） of the whole system, to the Centenaria. This writer mentions, however, a size larger than the centenaria, the centum vicenûm; and one smaller than the quinaria, the digitus rotundus, or pipe of exactly a digit in bore ( “Digitus rotundus habet diametri digitum unum; capit quinariae septuncem semunciam sextulam” ; a roundabout way of expressing 45/72 or 5/8, Frontin. 100.26 extr.). But the largest pipes now extant exceed 300 quinariae (Middleton, p. 466. For further details, see Frontinus, de Aquaed. 2463, with the notes of Polenus). Another mode of explaining the nomenclature was by the story that when Agrippa undertook the oversight of the aqueducts, finding the modulus inconveniently small, he enlarged it to five times its diameter, and hence the origin of the fistula quinaria (Frontin. 25). Of these accounts, that of Frontinus is the most generally accepted.
The subject of inscriptions on lead pipes was first noticed by Marini (Atti de' Fratelli Arvali, Rome, 1795); but it is much more fully treated in the magnificent work of the Commendatore Lanciani (I Comentarii di Frontino, Rome, 1880). Recent excavations in Rome have unearthed immense quantities of these inscribed pipes, many of which have unfortunately been melted down as old lead. From those which have been preserved, Comm. Lanciani has identified 81 sites of houses within the walls, 88 in the suburbs, and has gathered much valuable information as to the distribution of water from the various castella, and many collateral historical and topographical facts. The earliest existing inscriptions date from the reign of Augustus, the latest from that of Valentinian II., A.D. 375392: nearly all the intervening names are included. They are most numerous in the reigns of Domitian, Nerva, Trajan, Severus and Caracalla ; after Decius the number steadily declines. The facts usually recorded on them are the name of the reigning emperor, sometimes the consuls of the year, the Procurator Aquarum, the plumber who made the pipe, the owner of the house, the name of the estate, the fact of the water being an imperial concession; and many have numerals, showing the capacity of the pipe in quinariae (Middleton, p. 462; who adds a copious selection illustrating the above points, pp. 463465). Our specimen (fig. A) bears the inscription STATIONIS PROPRIAE PRIVATAE DOM. N. ALEXANDRI AUG.; indicating the private water supply of the imperial palace under Alexander Severus. Another, from the Palatine, has [DOM]VS AUGUSTANAE; while a pipe inscribed IMP. AUG. VESPASIANI . STATIO. VRBANA. AUG. belonged to the public waterworks of the city. Respecting the uses of pipes in the aqueducts, see AQUAEDUCTUS p. 154 a.
Of the earthen (terracotta) pipes we know very little. Pliny says that they are best when their thickness is two digits (1 1/2 inch), and that each pipe should have its end inserted in the next, and the joints should be cemented; but that leaden pipes should be used where the water rises. The earthen pipes were thought more wholesome than the leaden. (Plin. Nat. 31.57; Vitruv. l.c. § 10; Pallad. 9.11.) Water pipes were also made of leather (Plin. Nat. 5.128; Vitruv. l.c. § 8), and of wood (Pallad. l.c.), especially of the hollowed trunks of the pine, fir, and alder (Plin. Nat. 16.224).