Hypocaustum

vol. IX p.333-339

Hypocaustum

The Romans used this foreign word to refer to the system introduced in the 1st century BCE (see below) in Italy for underfloor heating in bathing and living rooms, which an extraordinary number of examples of are extant everywhere where there are Roman remains, especially in the north of the empire. <10> General points about the installation and use of hypocaust heating have already been discussed in the article on heating vol. VII p.2646ff. Here only a few particular things will be noted, without going into all the technical details for which a large number of diagrams would be necessary. The subject has been recently handled in detail with expertise and a knowledge of ancient sources and ruins by G. Fusch Über H.-Heizungen, Dissertat. der Techn. Hochschule Hannover 1910, with 30 panels. <20>

In Plin. ep. II 17, 11. 23. V 6, 23, hypocauston (in all the better mss.) describes the heating system structurally connected to the room to be heated (see vol. VII p.2651, 47ff.), in Stat. silv. I 5, 58 hypocausta (Plur.) refers to the whole heating system of a bath house, in Ulp. Dig. XXXII 55, 3 diaetae hypocaustae are the living rooms fitted out with hypocaust-heating, <30> cf. Epiphan. adv. haeres. II 52 p. 459B: ἐκκλησίας .. ἐν ὑποκαύστοις οἰκοδομεῖν, ὑποκαίειν τε κάτωθεν, ibid. D: ἐν ὑποκαύστῳ οἴκῳ τὴν δίαιταν εἶχον. In Vitruvius, who we owe the specific description of a hypocaust system to (see below), the word hypocaust doesn’t turn up. However, in V 10, 1 p. 125, 8ff. he calls the fireplace hypocausis, just as Plut. quaest. conv. III 10, 3 p. 658 e refers to the firing of a bath with ὑπόκαυσις (cf. Hesych. s. ἰπνοκήϊον .. οἱ δὲ τὴν ὑπόκαυσιν τοῦ ἰπνοῦ). <40> Finally, in an extract from Vitruv. de div. fabr. archit. 16 p.300, 18, Faventius uses hypocausteria for the heating stoves in a bathing complex.

The most important components of the hypocaust system are the following:

1. The oven, hypocausis (see above), according to Vitruv. loc. cit. p.125, 6 in Latin praefurnium. It was always outside the room to be heated, usually in a covered room, which was called propnigeon (Plin. ep. II 17, 11. Vitruv. p. 127, 19). <50> Whole buildings and larger rooms usually have multiple praefurnia. Conversely, Vitruv. p.125, 8ff. - whose accounts are confirmed by the Stabian and forum baths in Pompeii - requires that in double-baths, the hypocausis lie between the caldaria in the male and female sections, and should be shared by both. Usually, praefurnia are channels covered with brick or fire-resistant stone, which go right through the relevant outer wall into the cavity beneath the room to be heated (see below sect. 2). <60> In contrast, over the hypocausis in the baths, we find a large standing cylindrical stove or multiple linked stoves next to each other used to heat up the water, and the testudines alveolorum (Vitruv. 125, 13), the semi-cylindrical metal attachments to the pools (Mau Gött. gel. Nachr. 1896, 76ff.) are kept warm by the shared oven. <page break 333/334> But even here, the praefurnia are also connected to the cavity beneath the caldaria and tepidaria (an illustration of a hypocausis with an extant standing lead stove, testudo, and hypocaust from the villa in Boscoreale Mon. dei Lincei VII 447 and its front view Röm. Mitt. IX p.353). <10>

2. The hollow floor. The basic part of all hypocaust heating systems is the hollow room under the stone floors of the inner rooms or water pools which need to be heated used to circulate heat. According to the sources cited above, this part was specifically called ὑπόκαυστον, ὑπόκαυστα. The floors had to be kept up in some way or another. The technical terms for this are suspendere (Cic. Hortens. frg. 69: Sergius Orata primus balneola suspendit), <20> suspensio (Vitruv. 125, 18: flamma facilius pervagabitur sub suspensione), suspensura (Vitruv. ibid. 14: suspensurae caldariorum ita sunt faciendae), and a room or pool with a hollow floor was called pensilis (Plin. n. h. IX 168. XXVI 16, etc.), a word which has also gone over to mean ‘oven’ in Germanic languages (old high German pḗsle, middle high German pfiesel, Müller-Zarncke Mittelhochd. Wörterb. II 493. Kluge Etym. Wörterb. s. Stube). <30> The most common way of making it, evidenced by ruins, is given by Vitruv. loc. cit. p. 125, 14ff., cf. also Pallad. de re rustu. I 40, 2. The upper floor was supported by small pillars two or three feet high, which were built from quadrilateral or round brick plates at regular intervals, and which were covered with large plates extending over the middles of the pillars, and was made of a usually very thick screed material (up to ½m and beyond). <40> In place of small brick pillars, we also find pillars made of quarry stones or ashlars, of clay cylinders, etc., and the hollow spaces between them are sometimes replaced by some wide channels (cf. Jacobi Saalburg 245ff. Durm Baukunst der Etr. u. Römer² Abb. 395-398. Schleyer Bäder und Badeanstalten 55. Fusch Taf. IV and XI).

3. The hollow walls. In order to make even better use of the warmth produced by the hypocaust system, <50> the hollow spaces would also be connected to the walls and ceilings especially in the caldaria, tepidaria, and sudatoria in the baths. For this, people used either tiles with protrusions on the corners, tegulae mammatae, or rectangular (rarely round) clay pipes, tubi or tubuli, which, standing vertically, were set up over each other and next to each other, and were plastered and attached to the walls with cross-shaped nails and mortar (illustrations in Jacobi, Durm, Schleyer loc. cit., Fusch Taf. II), <60> cf. Seneca ep. II 90, 25: quaedam nostra demum prodisse memoria scimus, .. ut suspensuras balneorum et impressos parietibus tubos, per quos circumfunderetur calor, qui ima simul ac summa foveret aequaliter. Plin. ep. II 17, 9: transitu interiacente, qui suspensus et tubulatus conceptum vaporem .. huc illuc digerit. Proculus Dig. VIII 2 de servit. praed. urban. 13: non licet tubulos habere admotos ad parietem communem, .. quod per eos flamma torretur paries. <page break 334/335>

For any Latin speaker, these (and other) sources remove any doubt or alternative ideas about the purpose of the hypocaust, the suspensura, and tubulation. Because of this, when O. Krell tried to prove that the hollow floors and walls were instead used to keep things dry and that heating was actually done with coal pans in the work Altröm. Heizungen 1901, <10> he had failed from the get-go. His suggestions, which are often overlooked in scholarship, are also proved wrong by ruins (on this, cf. Blümner Berl. ph. Wochenschr. 1902, 398. Fabricius Korrespondenzbl. des Gesamtvereins 1903 p.211. Brauweiler Westd. Ztschr. XXIII 11. Durm 362. Wiegand Abh. Akad. Berl. 1908, 15 and also Fusch 15ff.). <20>

Heating up a hypocaust required a long time, and the use of the whole hypocaust system also required a lot of caution; in particular, the fire could only be moderate (Plin. ep. II 17, 26; cf. Stat. silv. I 5, 44. 58: languidus ignis). Through long heating channels between the praefurnium and the hypocaust, the flames were prevented from reaching the suspensura, and the heating gas was burned as completely as possible. For the most part, coal was used to create the heat. <30> However, the hypocaust systems, as the repair works still recognisable on almost all ruins show, were exposed to a lot of wear and tear. The question of how the combustion gases were got rid of hasn’t yet been fully solved. Where the hypocausts were connected to hollow walls, some of the tubuli may have been used to let off the combustion gases. More often, particular smoke outlets were attached to the walls and chimneys (Jacobi Über Schornsteinanlagen der Stabianer Thermen in v. Duhn and Jacobi Der griech. Tempel in Pompeji 30. Fusch 19ff. with fig. 23). <40> However, sometimes the air heated by the hypocaust was also directly lead into the rooms to be heated through closeable openings (see the literary sources vol. VII p.2651, evidence from the ruins Jacobi Saalburg 248. Schumacher Oberg.-raet Limes B 53, 18. Fusch 71ff.). <50> L. Jacobi has recently [as of 1914] reconstructed hypocaust heating systems at Saalburg, and used it to heat the museum and a room in the praetorium with great success (Durm 363. H. Jacobi Saalburg-Führer⁶ 60).

Following its name, the hypocaust heating system must be Greek in origin. In fact, the accounts about C. Sergius Orata (vol. II A p.1713, 48-50 [corrected from vol. II p.2650]), which can all be traced back to one famous speech of L. Crassus about fishery rights in the Lucrinian Sea, do not say that he used them to heat up water basins for his fishery. In the time of the Oscans, the baths at Pompeii didn’t have any hypocausts yet (Nissen Pomp. Stud. 144ff. Mau Pomp. Beitr. 137-141). At Pompey’s time, the luxury of floating baths (Plin. n. h. XXVI 16 iam primum pensili balinearum usu ad infinitum blandiente) had spread quickly into Rome and all Italy. <page break 335/336> This was when the caldaria at the Stabian baths got their hollow floors (Mau loc. cit. 143 thinks it was at around the time of Sulla). Tubulation in the walls was still unknown to Vitruvius, and in the baths at Pompeii it can also be observed that they are more recent than the installation of the hypocaust (Mau 142. 151; Pompeji in Leb. u. Kunst² Anhang 1913, 35). <10> Seneca is first aware of the nonsensical increase of the temperature in the baths, which was made possible at the time by drawing the heat through the walls as well (see above and ep. 86, 10). All the way through the empire, the hypocaust heating system was used by the the smallest complexes at the lime fortresses (the obergerm.-rät. Limes B 72 Taf. IV 2-4) up to the huge rooms and water pools in the imperial baths and basilicae in Rome and Trier. Today, it is still in use in Turkish and Arabic baths in the east (examples also in Schleyer 178ff.). <20>

[Fabricius.]