Aksum An African Civilisation of Late Antiquity Stuart Munro-Hay

were found too) which has close stylistic connexions with another found at Addi Galamo 
near Atsbi Dera (Caquot and Drewes 1955). The latter is associated with a plinth bearing 
an inscription which has been interpreted as reading `That he might grant a child to 
Yamanat (YMNT)' — (but see also Ryckmans 1958). Usually the Hawelti `throne' and 
statue are thought of as ex-voto offerings to the lord of the temple by richer citizens (de 
Contenson 1962); but Jacqueline Pirenne (1967) proposed that the two statues 
represented the divinity, and the covered thrones the naos, of each of the temples. She 
also suggested, with some likelihood, that some of the objects found at Gobochela, 
Hawelti and Enda Cherqos were actually older than the structures involved. They had 
been taken from a now-destroyed temple of the `Sabaean' period, either by descendants of 
the original dedicators of the altars, or by worshippers who still venerated the same gods 
but had lost the skill to produce such monuments. This would explain the juxtaposition of 
finely carved altar, inscriptions, thrones and statues with crude structures and rough 
pottery ex- voto objects. Hawelti also produced some bronze objects, rings and openwork 
plaques, similar to others found with pottery deposits in small pits at Sabea in the Agame 
district (Leclant and Miquel 1959). Numerous other articles came from the stele area and 
some apparently ritual deposits at Hawelti. The general impression is that the objects 
were left at the temple as reminders to the gods about a number of human affairs; 
childbirth or fertility in the family, crops and domestic animals, protection against wild 
animals, safety and prosperity of the house, and other such cares. Further possible ex-
voto objects consist of crescents and phallic or female figures from Sembel-Cuscet 
(Asmara) and Aksum (Tringali 1987).  
Both the temple at Yeha and the throne from Hawelti depict the ibex, the sacred animal 
connected with the worship of Ilmuqah. Perhaps the bull (also sacred to Ilmuqah) 
succeeded the ibex in popularity as an ex-voto offering in Ethiopia, since several bull-
figures have been found in Ilmuqah temples there, and one small schist image from 
Gobochela even has a dedication to Ilmuqah written on it (Leclant 1959: 50, pls. XXXIX-
XL; Drewes 1959: 95-7). The bull was also the symbol of Sin, the moon-god particularly 
venerated in the South Arabian kingdom of Hadhramawt, where the bull was depicted 
with the letters SYN on coins issued at the Hadrami royal castle called Shaqar in the 
capital, Shabwa. Most of these pre-Christian sites are marked by stelae, but it seems that 
they served as memorials or offering places rather than tomb- markers as later in Aksum. 
Apart from the disc and crescent symbols found on the altars, the later Matara and Anza 
stelae also bear these symbols, which may indicate that the great stelae at Aksum were 
similarly dedicated on the now- missing metal plaques at their summits. Pottery with the 
symbol has been found, and it appeared on the coins until the reign of Ezana, when the 
cross began to be used instead. The disc and crescent, however, presumably divested of 
its sacred character, continued to be used in Ethiopia as (apparently) a mint- mark on 
coins until the very end of the coinage (Munro-Hay 1984i: see Gersem, Armah).  
Apart from the `state' religions (if we may call them so) of the Sabaean or D`MT periods, 
there was probably an underlying stratum of more popular beliefs connected with 
animals, birds, and the various manifestations of nature, the weather, and so forth. It may 
be that some of these survived in magical rites connected with the kingship and enacted 

---

at the royal coronation, preserved by the continuance of that ceremony at Aksum in the 
time of the Solomonic restoration (see 
Ch. 7: 6
).  
A priesthood, presumably arranged in some sort of hierarchy, would have served the gods 
and made the offerings and sacrifices. It seems quite likely that the king, claiming divine 
descent, may have held a prominent place in it, perhaps as high-priest, at least of the 
dynastic deity Mahrem. However, there is no indication of this in the surviving texts, and 
nothing is known of the personnel of the pre-Aksumite temples. At the earliest period, the 
mukarribs of D`MT and Saba may have acted as both priest, offerer of sacrifices, and 
ruler, since these attributes are apparently represented in the meaning of the title 
(Ryckmans, J., 1951). Later, when the title mlkn (malik, king), and nagashi came into 
vogue, the greater part of the priestly side of the kingship may have been entrusted to one 
or several high-priests and their subordinates, but this is simply surmise.  
Paribeni (1907: 469-70) found near his `Ara del Sole', trenches with walls lined with 
stone (and in one case containing two rows of bricks), and filled with ashes. In the lower 
levels these contained no other material but ashes, and Paribeni suggested that they were 
dug to receive the material from animal cremations. He noted the care with which the 
pure ashes were deposited in the trenches, but added tha t there were no traces of 
carbonised bones. Around the Aksumite stelae, several deposits of carbonised bone were 
noted, and it seems very possible that dedicatory meals were prepared as part of the 
ceremonies of burial during the pagan period.  
2. The Conversion to Christianity 
 
The primary evidence for the conversion of Ethiopia in the reign of Ezana in the fourth 
century is found in the king's own inscriptions and coins. In the former (
Ch. 11: 5
), the 
locutions used to express his devotion to the gods are altered to Christian forms. The 
coins also abandon the disc and crescent symbol and replace them with a cross or several 
crosses, and a cross is even found on one of Ezana's inscriptions written in the Epigraphic 
South Arabian script, on the reverse of a Greek text which opens with Christian 
phraseology (Schneider 1976ii). An important feature of the coinage, already briefly 
noted (
Ch. 9: 3
), is that Ezana's Christian issue in gold with the name written Ezanas, is 
of the weight in use before the reform of the Roman currency by Constantine the Great in 
324AD. His next issue, on which the name is written as Ezana, followed the new pattern. 
This means that at some time relatively close to 324, Ezana had already decided to 
proclaim his new faith on his coinage. Even if we imagine that coins of the earlier weight 
might have been issued at Aksum for a few years after Constantine's reform, we still have 
a very early date for the conversion of Ezana and the appearance of the cross on 
Aksumite coinage (Munro-Hay 1990).  
This `official' conversion of the king is confirmed by Rufinus (ed. Migne 1849: 478-80), 
a contemporary Latin writer, who derived his information from Aedesius of Tyre, who 
had been a prisoner and servant in the royal household at Aksum with Frumentius, the 

---