**1. Introduction**

The provision of daylight in heritage buildings is rarely considered adequate in terms of responding to contemporary standards issued by international institutions and which have been wrongly based on Daylight Factor (DF) or climate-based daylight modelling (CBDM) [1]. Therefore, daylighting levels afforded by the original daylighting strategy in heritage buildings are rarely maintained as an intrinsic heritage experiential value in the rehabilitation or re-use of these buildings. In the context of public bathhouses (hammams), daylight provision is frequently done from the roof space. A single oculus at the centre of a dome is a feature that is commonly found in the architecture of large Roman public bathhouses as illustrated in Pompeii's Roman baths [2].

The decline of the Roman and Byzantine public bath institution in the West was followed by the development and proliferation, during the early Islamic era, of small public baths, reminiscent of the Roman baths known as balnea. These baths feature three successive bathing rooms of increasing temperature, as well as the underfloor heating system known as hypocaust, both were maintained in the architecture of Islamic hammams. However, daylighting provision was replaced by a number of small oculi, pierced in the vaults and domes of the hammam roof, becoming a distinctive characteristic of hammam buildings. This architectural feature can be seen in two early Islamic baths: Hammam Qusayr Amra, an 8th Century structure [3] in the Jordanian desert (Figure 1) and in Hammam Aghmat, an 11th Century Almoravid structure [4] excavated in Aghmat, 30 kilometres South East of Marrakech in Morocco [4,5]. The Aghmat hammam [6] displays rows of circular openings or oculi in the three vaults covering the cold, warm and hot rooms (Figure 2)*.* Such a configuration will continue unchanged in the architecture of Moroccan and Andalusian hammams for many centuries onwards [7–9]. The small circular piercings of 18 to 25 cm in diameter, located in the vaults and domes of the hammam roof are capped with blown glass bells. These allow shafts of daylight and sunlight to penetrate into the bathing spaces, cutting through the thick steam and creating different atmospheres at different times of the day. They constitute one of the unique and authentic experiential qualities of hammam buildings as illustrated in the case of one Mamluk and one Ottoman hammam (Figure 3a,b). Known in Arabic as little moons "Qamariyyat" or little suns "Shamsiyyat", they make a clear connection between the bather and the changing sunlight conditions of the sky. They also allow low levels of daylight into the steamy bathing spaces, hence maintaining some level of visual privacy, bearing in mind the strict religious codes governing nudity and gazing at another bathers' body. The act of washing one's body is indeed not a task that requires high levels of daylight, however, the manipulation of daylight in hammam spaces is believed to be highly linked to cultural and spiritual norms. It is argued in this paper that the vernacular daylighting system in hammams constitutes an intrinsic part of the heritage value and authenticity of this building type that needs to be fully understood in order to be properly rehabilitated in existing heritage structures and carefully reintroduced in the architecture of contemporary hammams in Morocco and beyond.

**Figure 1.** Qusayr Amra three vaulted bathing rooms with small oculi [2].

**Figure 2.** Hammam Ahgmat: 11th Century Almoravide era, Morocco [3].

**Figure 3.** Hammam lightscape in the hot room of (**a**) Hammam al Hussayniya in Cairo (Mamluk era); (**b**) Hammam al Bacha in Acre (18th Century Ottoman era).

Existing studies on hammam daylighting systems are few and far apart. The most recent studies focus on Ottoman hammams that have lost their original function. This is the case of the study conducted by Al Maiyah et al. [10] which rightly argue the importance of understanding the daylighting system in vernacular buildings, as this is rarely considered as a key feature worth preserving as part of adaptive reuse strategies of heritage buildings. The study focuses on how to maintain the original daylighting conditions in the hot room of the Demirci hammam in Bursa, while allowing it to function as a museum or an art gallery; the hot room being the only surviving part of this Ottoman structure. The main concern is to establish how to reuse the existing vernacular daylighting system without exposing the exhibited artefacts to levels of light that are likely to be damaging. Measurements were made to validate a simulation modelling tool, Radiance, used with a digital model of the building produced using Integrated Environmental Solutions (IES) software. The aim was to develop an understanding of the behaviour of daylight in this Ottoman hammam hot room during a whole year and determine the most appropriate spaces for placing museum or art gallery exhibits. The study highlights that the average monthly illuminance level on the South wall of the hot room remains under 140 lx in the summer months in Bursa-Turkey (Latitude: 40◦1144" N Longitude: 29◦0336"). Although this study rightly argues the need to avoid relying solely on lighting standards for the adaptive reuse of heritage buildings, and work with the existing lighting qualities of the spaces in a vernacular hammam, it does not provide an understanding of the nature of the hammam vernacular daylighting system in all the bathing spaces as it concentrates on analysing one main redundant space.

Another study focuses on investigating the vernacular daylight provision in seven still surviving Ottoman hammams in the city of Thessaloniki in Greece [11]. Measurements of daylight levels were, however, conducted in one case study building only, the Bey Hammam, a 15th Century twin structure used as a cultural centre was selected for recording daylight levels under different sky conditions in March 2008. The results of these measurements are presented graphically on the plan of the building with illumination levels plotted on a grey scale, ranging from 0 lx to 120 lx at a step of 10 lx [10]. Measurements were made using a lux meter at different points of the three bathing spaces and were carried out during a single day. This study indicates a clear correlation between the intensity of daylight, the type of activities carried out in the spaces and the level of visual privacy they require. The central spaces under the domes received the most light (due to the large number of oculi found in the domes of Ottoman hammams). This is the case of the central marble table in the hot room (the main shared space in Ottoman baths), whereas the peripheral spaces used for individual washing have much lower levels of daylight, indicating a clear correlation between the intensity of daylight and the level of privacy required by the bathers.

It is clear that the al Maiyah [10] and Tsikadoulaki et al. [11] studies focused on a single case study of Ottoman hammams that have lost their original function: One in Bursa, (Turkey) the other one in Thessaloniki (Greece). In both studies, daylight measurements, limited in time and space, were made in redundant structures without the steam conditions typical of a working hammam. Furthermore, all case study buildings where daylight levels were measured focus on a single bathing space, usually the hot room. However, it is clear from these studies that daylight levels in Ottoman hammam bathing spaces are relatively low and are in most cases below 140 lx as one moves away from the central domed part of the hot room.

This low level of daylighting in hammam bathing spaces is also confirmed by an earlier study carried out by Mahdawi and Orehounig [12,13] in the context of an EU-funded research project (HAMMAM 2008). Mahdawi and Orehounig collected data on indoor environmental (thermal) conditions and outdoor microclimatic conditions in the immediate vicinity of one traditional hammam in each of Egypt, Turkey, Morocco, Syria, and Algeria over a period of one year. Horizontal illuminance was measured (at one metre from the floor) at a single point in the centre of the different bathing spaces of the five case-study hammams of Cairo, Ankara, Fez, Damascus and Constantine. These measurements are of indicative character only and not reliable as they were carried out at different seasons in each of the five hammams and under different sky conditions in different geographies. Furthermore, measurements were made in conditions where the vernacular daylighting system was either in a poor state of repair or completely redundant and did not exclude the contribution of electric lighting at the time of the measurements. However, the results of horizontal illuminance measurements clearly indicate consistently low levels of lighting (mostly below 100 lx) in the five case study buildings of different historic eras and geographical locations. These measurements present however a number of limitations as they do not link to outdoor sky conditions at the time of the measurements and are not directly comparable. Furthermore, they fail to convey a clear understanding of the hammam daylighting levels afforded by the vernacular oculi system. Despite their limitations, it is clear from these previous studies that daylight levels in the bathing spaces of hammam case-studies located in different geographies (north and south of the Mediterranean) are generally low, varying between 50 and 150 lx.

This literature review has revealed that there have been no studies so far that attempt to develop an understanding of daylight levels in all the bathing spaces of heritage hammams under working steam conditions. Measurements of daylight levels in working heritage hammams where the vernacular daylighting system is still in operation, or has been fully restored, are completely non-existent. The lack of such studies makes it difficult to specify the vernacular hammam natural lighting for hammam rehabilitation purposes and for the design of new built structures that aim to create an authentic daylighting hammam experience. This is needed in all of the Maghreb countries of North Africa where the hammam tradition is still alive. Morocco is where the largest number of working heritage hammams are found and where housing planning regulations dictates the inclusion of a hammam facility in every new residential neighbourhood [14]. It is estimated by the Moroccan federation of hammam managers, that Morocco has more than 12,000 operating hammams, however, it is very likely that the number is much higher as there have never been a systematic national census for hammams. These operate day and night and tend to rely on electric lighting during daylight hours because of their redundant vernacular daylighting system or their poorly designed daylight provision. No studies have been carried out so far to develop an understanding of the hammam vernacular daylighting strategy and the unique spatial experiential qualities it provides. This paper provides a timely and much needed historical and architectural understanding of the hammam vernacular daylighting strategies that embed tacit cultural and social norms for visual privacy and spiritual connection to the sky. In order to do so, the following methodology was adopted.

### **2. Materials and Methods**

A representative sample of heritage hammams from different historical periods was selected in the old city of Marrakech. A total of 13 still functioning heritage hammams out of a total of fifteen were surveyed and recorded for the first time by the author, allowing the production of their plans, sections and elevations, as well as their roof plans and a photographic record of their various spaces (see Table 1 and Figures 4–6). More recently built hammams were deliberately excluded as the focus of this study is on the original vernacular daylighting system in heritage structures. The examined sample of hammams represents 90% of the total number of heritage hammams, located in every residential neighbourhood, within the proximity of small or large mosques inside the UNESCO world heritage intra-muros urban fabric of the city of Marrakech.


Hammams 13 and 14, bolded in Table 1, have not been surveyed as they were transformed into spas.

In order to establish a clear understanding of the hammam vernacular daylighting system, the following methodological steps have been followed:

*Step 1*: Architectural analyses of heritage hammam roof plans with their oculi number, location and configuration

The plans and roof openings of the 13 hammams, documented during field work in Marrakech, are systematically analysed to establish whether there are recurrent patterns in the number, location and configuration of roof openings (oculi). The aim of such analyses is to identify any tacit rules that are applied for the configuration of the Moroccan hammam vernacular daylighting strategy.

*Step 2*: Analyses of ratios of total roof openings area over internal floor area in each the three bathing spaces of the 13 surveyed hammams

The calculation of the percentage of the total area of roof openings for each bathing space in relation to its internal floor area will reveal whether there are any recurrent ratios and whether there any variations between the cold, warm and hot rooms. This is achieved through the comparison of these ratios in the 13 investigated historic structures. This analysis also allows for the extraction of any underlying tacit rules that can inform both the rehabilitation of existing heritage structures and the design of future new hammams.

*Step 3*: Measurement of horizontal illuminance afforded by the vernacular daylighting system after its rehabilitation in a working heritage hammam

A working heritage hammam, presenting a vernacular lighting configuration of roof oculi over its bathing spaces, has been selected in order to restore the blown glass caps that originally covered the roof openings and carry out measurements of daylight levels afforded by the system in each of the three bathing spaces under working conditions. Water resistant HOBO pendant data loggers for temperature (range of −20 ◦C to 70 ◦C) and light (range of 0 to 320,000 lx) were chosen due to their discreet small size and their ease of installation in working hammam spaces with high levels of humidity [15].

They were installed at similar and comparable locations in each of the three retrofitted bathing spaces (of similar size, colour and texture and hence similar reflectance) and on the roofs to record outdoor horizontal illuminance under open sky conditions. The data loggers were placed on the south facing parallel walls of each of the three bathing spaces and at a height of two metres from the floor (away from the bathers), allowing for comparable points of measurements. All data loggers were synchronised to measure horizontal illuminance continuously every 20 min in July, August and September of 2016. This allowed for three measurements per hour for three months which is sufficient to establish the levels of horizontal illuminance falling at similar points in each of the three similar hammam bathing spaces during the summer season of 2016.

**Figure 4.** Location of the surveyed heritage hammams of Marrakech with numbers corresponding to the names of the hammams in Table 1.
