top of page
acroteria render.jpg


May to June 2020

The time when final designs were needed for a set of acroteria fortuitously coincided with finding a sculptor who was well-placed to collaborate with the stoneyard. With a contest between craft and technology in mind, we set about creating a digital 3D model so as to generate both 1:1 line drawings for the sculptor’s clay prototype and files suitable for CNC cutting. This would give us the option of a robot-cut alternative prototype too.

Design Evolution


1) Idea and Premises


Given the long roofs of the bath complex we favoured breaking up the skyline of with acroteria (from the Greek akros-: topmost, end-most). At different times of day such devices can catch the sun, become silhouetted at twilight, or be picked out by spotlights for evening gatherings. The familiar palmette-on-volutes composition presented an obvious starting point; we use palmettes elsewhere in the project, there are palm trees round about, while volutes could evoke waves of the sea below. Initially we had no intention to treat these unconventionally; only when details were not quite working out did the idea of up-turning them come into play.

2) Design Development


In line with our preferred way of working, we kicked off with hand-sketches without looking for precedents in books or the internet, first experimenting with sketches and elevations at 1:5, and then one half at 1:2 which we could mirror. Next came 3D modelling and triangulation between manipulating the 3D model on screen, printing out views and projections, and revising them by hand. At this stage researching how details had been handled in the past became extremely useful – it would be mad not to learn from a history and all the clues for improvement it can offer.


The palmette

Due to the leisure character of the building and the real palms in the natural environment we wanted the palmettes leaves / fronds to look more organic than some 19th- and 20th-century examples. In tune with the way we have treated acanthus elsewhere in the project the important detail is to V-cut the individual fronds so as to catch shadows and augment a sense of volumetric plasticity. The only technical challenge was to find an effective way of achieving this with the software used. 


The volutes

In both antiquity and the Greek Revival period, some designers supported the palmettes with paired rising volutes, comparable to those of the so-called Aeolic capital. Others employed flowing S-scrolls that were a staple in antiquity for linking palmette and lotus motifs. The fact that there never developed a canonic solution may indicate a certain level of difficulty in arriving at a truly satisfying outcome. 

acroterion MWJ sketches.jpg

early sketch

sketch showing flipped ionic volutes

Indeed, play as we might with the “Aeolic approach”, we couldn’t find a pleasing way to terminate the volutes /scrolls at the sides. Besides, even if this was sometimes done in antiquity, cutting off the S-shape of lotus-and-palmette motifs seemed inherently unsatisfactory. On the other hand a complete S-shape generated too much height relative to the overall elevation. (Our hands were tied by having fixed the width of the base blocks, which had already been fabricated integrally with the cornices.) In part out of frustration, in part out of curiosity, we tried flipping the volutes into an upside-down Ionic arrangement (see sketch with flipped Ionic volutes). This meant completely over-hauling the earlier 1:2 elevation to produce a new version. Once convinced it could work we moved back to revise the 3D model, testing the effects from all angles.


In the absence of a historically established instance, flipping appealed to us as an interesting avenue to explore. At the same time confidence it could work was provided by knowledge of comparable flips in ancient Greece either side of 600 BC. This was a time of intense creativity when previous habits of building were revolutionised and Doric and Ionic style began to crystallize. Metalworkers, vase painters and craftsmen freely inverted, rotated, mirrored and flipped elements such as volutes in the pursuit of aesthetically pleasing effects. [MWJ 2014, 163-167] (pictures). Architects were open to such possibilities too, even if more constrained by materials, technique and conventions. Indeed, the creation of the Ionic capital may have been triggered by flipping rising-volute motifs of the kind that had widespread currency all over the Mediterranean [MWJ 2014, 126, 214-215] (pictures) Up-turned scrolls also feature on a number of Greek and Roman altars and sarcophagi.

3. Further historical notes

Possibly inspired by antecedents from further East, the Greeks initially used acroteria to decorate the pedimented temple roofs they created having invented the roof tile in the 7th century BC. Some early temples had a prominent central acroterion, and some of these were shaped as a disc though a single palmette rising out of volutes went on to be popular. The lateral acroteria tended to be smaller and often took the form of a sphinx. Experimentation with comparable terminations also occurred in the context of marble stele or funeral markers, as well as decorative terracotta roof edging. 


The classic grouping of a central palmette acroterion flanked by half-palmettes for the corner acroteria was used before the Parthenon, but here it flowered into a gorgeously elegant celebration of virtuosity. The supporting foliage and the fingers of the palmettes were carved into daring filigree open-work. (Not surprisingly only fragments survive.) Less exacting variations on this theme became popular on sacred buildings and tombs down to the Christian period, when for the most part they fell out of use.


In the medieval and Renaissance periods the skyline of prestigious buildings could be populated by all kinds of things – statues, gargoyles, crosses, finials, pinnacles, baldachins, along with real or fictive chimneys and crenellations. Statues became particularly popular from the time of Sansovino, Sanmichele and Palladio. Ancient types of acroteria were almost absent, however, even into the early Greek Revival period. They eventually re-appeared only on the back of archaeological discoveries. Soane, who owned plaster casts of Roman examples, used palmettes to decorate rooflines, while Alexander “Greek” Thomson returned to the archetypal formula of one central and two flanking palmette acroteria on his Caledonia Road Church and St. Vincent Street Church, both in Glasgow.





Danner, Peter 1989 > Griechische Akrotere der archaischen und klassischen Zeit (Rivista di Archeologia, supp. 5).


Delivorrias, Angelos 1974 > Attische Giebelskulpturen und Akrotere des fünften Jahrhunderts.

Goldberg, Marilyn Y. 1982 > Archaic Greek Acroteria, AJA 86: 193-217.


Gropengieβer, Hildegund 1961 > Die pflanlichen Akrotere klassicher Tempel.


Himmelmann, Nikolaus 2005 > Grundlagen der griechischen Pflanzendarstellung, Nordrhein-Westfälische Akademie der Wissenschaften, Vorträge G 393.


Ira S. Mark 1977 > Hesperia: The Journal of the American School of Classical Studies at Athens, The American School of Classical Studies at Athens 

Meurer, M. 1896 > Das griechische Akanthusornament und seine natürlichen Vorbilder, JdI 11: 117-159.    


Rykwert, Joseph 1994 > On the Palmette, Res. Anthropology and Aesthetics 26: 11-21.


---- 1996 > The Dancing Column. On Order in Architecture.

Wilson Jones, Mark 2000 > Origins of Classical Architecture.


bottom of page