DUFFY, JOHN, A Small Box in John Moschus , Greek, Roman and Byzantine Studies, 24:1 
(1983:Spring) p.93 

A Small Box in John Moschus 

John Duffy and Gary Vikan 

I 

JOHN MOSCHUS (ca 550-634) was a monk and spiritual writer 
whose one surviving work, the Pratum Spirituale, 1 became a very 
popular manual of devotional reading, especially in monasteries, 

throughout the Byzantine world. Composed probably in the second 
decade of the seventh century, the work consists of a long series of 
edifying stories about monks and their lives in many parts of the 
Near East, and it provides a wealth of information on religious and 
social conditions in that part of the Empire during the late sixth and 
early seventh centuries. 

One of the distinguishing features of the Pratum is its language~ 
Moschus writes Greek that tends towards the colloquial, and he is 
one of a handful of authors who bring us close to the Byzantine 
vernacular of the period.2 This informal style allows him to admit 
certain words from everyday language which, if writing in a more 
elevated manner, he would likely avoid. The present note concerns 
one such term which will be shown to occur in two passages of the 
Pratum and whose meaning, despite our ignorance of the etymology, 
can be established with more confidence than heretofore. 

In ch. 203 (3093A-C Migne) we are told the story of a gem-cutter 
who embarks on a business trip by boat, accompanied by his chil­
dren, taking with him precious stones and pearls. During the voyage 
he is warned that the sailors are plotting to throw him overboard in 
order to get the gems~ whereupon he commands his children to fetch 
the stones and toss them into the sea, thereby frustrating the evil 
intentions of the crew. According to the Greek text, the man orders 
the children to fetch the gems by saying cpePE'TE T£l /-WV 'tKia. An 

I Also known as the Leimon or Leimonarion or Neos Paradeisos. The text is in 
Migne, PG 87(3).2852-3112; a new edition is expected from Dr P. Pattenden. For a 
good recent account of Moschus and his work see H. Chadwick, "John Moschus and 
his Friend Sophronius the Sophist," JThS 25 (I974) 41-74, reprinted in his History and 
Thought in the Early Church (London 1982). 

2 See R. Browning, "The Language of Byzantine Literature," in S. Vryonis, ed., The 
Past in Medieval and Modern Greek Culture (Malibu 1978) 112. 

93 



DUFFY, JOHN, A Small Box in John Moschus , Greek, Roman and Byzantine Studies, 24:1 
(1983:Spring) p.93 

94 A SMALL BOX IN JOHN MOSCHUS 

accompanying note in Migne3 explains that the word 'I.KLa is unat­
tested and that, to judge by the Latin version capsulas, 4 it may be a 
mistake for f3I.Kla.5 The attempt at emendation is admirable, in view 
of the context, but it is not the correct one. 

A more attentive ear to the man's words would detect something 
strange. Would a Greek speaker, especially in the colloquial atmos­
phere favored by Moschus, be likely to say cPEPETE TeX J.LOV 'I.KLa? 
Hardly: then as now one would expect cPEPETE Ttl 'I.KLa J.LOV. The text, 
therefore, is faulty. But the problem is not deep-seated, for it can be 
removed by a slight shift of the elements, namely by reading cPEPETE 
Ttl J.LOv,iKux.6 This again confronts us with a rare word, J.LOV,iKWV~ but 
fortunately it is attested twice in another story of the same John 
Moschus, and apart from that in no other Greek author. The term 
appears in ch. 79 of the Pratum: 

1TapaYEVO/-UVOL Ei<; l:Ei\EVKELaV rY,v 1TPO<; 'AVTLOXELaV, (T1)VErUX-

0/-UV ~ a{3f3/!. 9E08WPct' E1TI.O'KOmp rij<; avrij<; l:Ei\EvKELa<; 1Toi\EW<; 
Kat 8nrY7J(TaTO ,y,piv i\EYWV OTL, "fEYOVEV E1Tt TOV /UlKapWV dW­
VV(TWV TOV 1TPO Ep..oV E1TL(TK01TOV TaVTT1<; rij<; 1TOi\EW<; TL TOWVTOV' 
~ , \ \ , "'8 ' 'i\ IJ \ \ i\ ' T/v TL<; aVT/p 1TpaY/UlTEvTT1<; EVTav a 1Tavv EV atJT/<; Kat 1T OV(TW<;' 
~ '" \. \ '" ~ IJ' "'-' " '" \ \ T/V oE aLpETLKO<; TWV L.EtJT/POV uuY/UlTWV, EXWV oE 1TL(TTLKOV KOLVW-
VOl)VTa rfi ay0 Ka8oi\LK'fi Kat a1TO(TToi\LK'fi EKKi\T/(T0. OVTO<; 0 1TI.(T­

TLKO<; KaTel TO [80<; T71<; xwpa<; Ei\a{3Ev KOLVWVLaV rfi ay0 1TE#L7TT'{I, 
\ f3a\.}~ '\!I r"'8 ' '""~" , KaL I\.WV aVTT1v EV p..oV."LKLct', a1TE ETO EV Tct' tuLct' ap/UlPLct'. (T1)V-

, ",.1 \ \ ., , \ \ """8'" , K E{3T/ Ut /-UTa TO aywv 1Ta(Txa TOV 1TL(TTLKOV 1TEfoAAfI T/vaL EV WV-
(TTaVTLVOV1TOi\EI. 8w 1TpaY/UlTELa~, Ea(TaVTa KaTel i\7J8T/V Tel<; aYLa<; 
KOLVWVLa<; EV T~ ap/UlpUy aVTOl). 8E8wKEV 8E TO Ki\Ec.8wv TOl) ap/Ul-

, '" '8' , """"!i: ", t'8' PWV Tct' av EVm. EV J.Ul!- ovv av0"Sa<; TO ap/Ulpwv ° av EVTT/<; 
"" , y' " ,~, , " EVPEV TO p..oV."LKWV EXOV Ta<; aYLa<; KOLVWVLa<; ... 

After arriving at Seleucia near Antioch we met with abbot Theo­
dore, the bishop of the same Seleucia, who told us a story, saying, 
"The following happened in the time of blessed Dionysius my 
predecessor as bishop of this city. There was a certain businessman 
here, very devout and rich, who belonged to the heresy of Se­
verus; he had an agent who was a member of the holy catholic and 
apostolic church. This agent, following the local custom, would 

3 The note is taken over from the seventeenth-century edition of Moschus by 1. B. 
Cotelier. 

4 The Latin translation printed in Migne is the work of the fifteenth-century human­
ist Ambrogio Traversari. 

6 That is, a diminutive of 0 {fiKO<; (jar, cask). Lampe, Patristic Greek Lexicon s. v. 
'LKWIl, evidently accepts this suggestion, listing the one occurrence as an "error for 
{3ucWIl, vessel, box." 

6 We should point out that Migne tacitly changed the accent in ,LKia; Cotelier had 
printed ,iKta. 



DUFFY, JOHN, A Small Box in John Moschus , Greek, Roman and Byzantine Studies, 24:1 
(1983:Spring) p.93 

JOHN DUFFY AND GARY VIKAN 

take communion on Holy Thursday. So having placed it in a mouzi­
kion, he stored it in his own cabinet. Now it happened that after 
Easter the agent was sent to Constantinople on business, forgetting 
that he had left the holy communion in his cabinet. He had given 
the key of the cabinet to his employer, and one day the employer 
opened the cabinet and found the mouzikion containing the holy 
communion ... "7 

95 

What is the meaning of J.WV~iKtOv in this passage? Du Cange8 

suggested that it may be a piece of cloth or napkin, and there is 
nothing in the description that would make that impossible. In put­
ting forward the idea Du Cange may well have been following the 
lead of Traversari, who rendered {3aAWV av'T-Y,v EV J.WV~tKicp by involvit 
in linteo mundissirno (where the adjective is interpretative) and EVPEV 
'TO J.Wv'UctOV by invenit Unteolurn.9 Stephanus has no entry for the 
word. Sophocles does, but ventures no explanation. Lampe is more 
definite, describing it as a "box inlaid with mosaic. "10 

Now that we have recognized a second instance of the word, we 
are perhaps in a better position to evaluate it. We can say straight­
away that the possibility of a cloth (Traversari, Du Cange) is more or 
less excluded by one detail in the gem-cutter's story: before the 
children bring the J.Wv,Ucta, the father spreads a piece of linen (&7TAOL 
<Ttv8ova)~ when the J.WV~iKta arrive, he opens them (avo~a~) and 
proceeds to layout the precious stones before he has them tossed 
into the sea. If the stones were already in cloth, would another be 
necessary? We should note too the different verbs used for 'open­
ing'. Lampe's definition, on the other hand, at least in so far as he 
opts for a box, is strengthened by the new passage. From the point of 
view of security, boxes would seem much more appropriate for gems. 
And, to return to the second story, while linen could be used for 
keeping the eucharist, a box would make at least as good sense. 
Lampe's added "inlaid with mosaic," however, would seem to de­
pend only on the assumption that J.WV~iKtOV is derived from the term 
for mosaic work (J.WV<TELOV or J.WV<Twv). This is a tempting idea, but 

7 2936C-D Migne; only the relevant first half of the story is quoted here. The phrase 
Eha{3Ev KOWWViaV rfi uy0 7TEJ..I.-7TT'"[/ could conceivably mean "he would take (sc. home) 
communion (sc. to have) for Holy Thursday." Whatever the precise meaning of the 
slightly ambiguous Greek, the second part of the account makes clear that the agent 
took home the eucharist in order to have it on hand for Holy Thursday of the following 
year. 

8 Glossarium s. v. (placed after the entry Mowa). 
9 Du Cange in his entry ,iKWV specifically mentions Traversari's capsulas. 
10 Stephanus, Thesaurus Graecae Linguae~ Sophocles, Greek Lexicon of the Roman and 

Byzantine Periods; Lampe, A Patristic Greek Lexicon. 



DUFFY, JOHN, A Small Box in John Moschus , Greek, Roman and Byzantine Studies, 24:1 
(1983:Spring) p.93 

96 A SMALL BOX IN JOHN MOSCHUS 

it ignores the linguistic problem created by the letter zeta. The ety­
mology must remain an open question. 

We may conclude, then, from the two surviving examples that at 
the very least /-WV~LKto" denotes a small box for keeping precious 
objects. 

II 

Do these passages in Moschus help explain any genre of 'small box 
for keeping precious objects' extant from the late antique period?ll 
The obvious candidate for identification with Moschus' /-WV~LKto" is 
the ivory pyxis.12 For decades it has been supposed that those pyxides 
with Christian iconography were intended for eucharistic or eulogia 
bread; 13 ch. 79 of the Pratum offers important new evidence from a 
region and period of the pyxides' manufacture suggesting that there 
was a demand for such receptacles and indicating plausible circum­
stances in which they might have been used.14 Two further points are 
noteworthy. (1) While most scholars have assumed that ivory pyxides 
were altar implements, Moschus places his host box within the realm 
of private piety; this might explain why pyxides do not appear among 

II Archaeology provides interesting leads but no answers for the etymology of J.Wv~i­
KWIJ. Linguistic problems aside, Lampe's 'box inlaid with mosaic' is unsatisfactory 
because no such genre is known from the period. On the other hand, the possibility of 
some connection with the Muses is inviting, for two reasons: (1) at least one well­
known genre of late antique box, the so-called Pannonian caskets, frequently bear 
bronze relief images of the Muses; (2) Muses, authors, and scroll boxes (the latter 
usually called capsae or capsulae) are frequently associated in late antique art. A small 
re-adapted version of such a scroll box, perhaps once decorated with the Muses, might 
have been called a J.WV~iKWII. Yet here too the problem of zeta applies. For the Pan­
nonian caskets see H. Buschhausen, Die spiitromischen Metallscrinia und frilhchristlichen 
Reliquiare I (Vienna 1971) 9ff nos. A7, 23, 25, 26, 38, 40, 41, 45. For some Muses, 
authors, and scroll boxes, see K. Weitzmann, ed., Age of Spirituality: Late Antique and 
Early Christian Art (New York 1979) nos. 238-42. 

12 W. F. Volbach, E/fenbeinarbeiten der Spiitantiken und friihen Mittelalters 3 (Mainz 
1976) 69ff, 103ff. About one third of the nearly five dozen specimens comprising this 
series of cylindrical ivory boxes bear pagan themes, the remainder Christian; traditional 
datings vary from the fourth to the seventh century, while localizations range over a 
variety of Mediterranean centers from Italy eastward. Neither Du Cange, Lampe, nor 
Sophocles provides an attestation of the word mJ~t~ which can be directly linked to 
Volbach's ivory boxes. For the antique use of the word see H. Graeven, "Pyxide en 
os," MonPiot 6 (I 899) 161 ff. 

13 See most recently V. Elbern, "Altar Implements and Liturgical Objects," in Weitz­
man (supra n.ll) 596f and no. 549; and A. St Clair, "The Visit to the Tomb: Narrative 
and Liturgy on Three Early Christian Pyxides," Gesta 18.l (I979) 132. 

14 For evidence more distant in time and place see H. Leclercq, DACL 14.2 (1948) 
1983ff s. v. "Pyxide "; also Du Cange s. v. 7TV~ioll. 



DUFFY, JOHN, A Small Box in John Moschus , Greek, Roman and Byzantine Studies, 24:1 
(1983:Spring) p.93 

JOHN DUFFY AND GARY VIKAN 97 

the objects portrayed in early Byzantine representation of the divine 
liturgy.15 (2) For Moschus, consecrated bread was worthy of the same 
protection as gems; this might explain why so many pyxides (espe­
cially those with Christian iconography) bear locks.16 

It is reasonable to suppose that some pagan ivory pyxides (e.g. 
those with scenes such as Venus and Adonis, Volbach no. 98) may 
have been used for gems,17 and that some Christian pyxides (e.g. 
those with the Blessing of the Loaves and Fishes, nos. 165, 166) may 
have been used for the hosP8 It is quite another thing, however, to 
suppose that the genre they represent was developed specifically in 
response to those functions-that the ivory pyxis was, depending on 
its iconography, a jewelry box19 or a box for holy bread. 

The world of John Moschus knew several distinctive types of small 
boxes of idiosyncratic design and function. There were, for example, 
pen boxes with trays for reeds and wells for ink, medicine boxes with 
images of healing deities on their sliding lids and tiny compartments 
for various remedies, relic boxes in the shape of small sarcophagi, and 
weight boxes with rows of variously proportioned holes suitable only 
for a specific set of weights.20 That Moschus assigned a single name to 
objects with two such different functions and contexts strongly implies 
that his J.LOV~iKtOJ) was a multi-functional, 'generic' box. The same was 
probably true of the ivory pyxis, for in addition to their possible use 
for consecrated bread and for gems, there is evidence suggesting that 
they served also as containers for incense and for offerings: 

Incense. In early Byzantine art Sts Stephen and Zachariah, and 
occasionally the Women at the Tomb, hold a censer in one hand and 
a small domed cylindrical box in the other; such a box, though with­
out a cover, is held by the sacrificing woman on the ivory diptych 
panel of the Symmachi.21 Not only do ivory pyxides match these 

15 As on the Riha and Stuma patens, and in the Rossano Gospels. Rather, it is the 
paten that is typically used for the host. 

16 See Volbach no. 193a for a Christian pyxis with its original lock intact. The box of 
the Seleucian Christian seems not, however, to have had a lock, as his employer 
needed only a key to the cabinet in which the ~v,i,KwlJ was kept in order to gain 
access to the consecrated bread. 

17 See Graeven (supra n.12) 163. 
18 See L. Kotzsche in Weitzmann (supra n.ll) no. 549. 
19 Jewelry boxes appear with some frequency in late antique art, and almost invaria­

bly they take the form of a rather large rectangular chest with a hinged lid. See Weitz­
mann (supra n.ll) figs. 32, 34, 35. 

20 See W. M. F. Petrie, Objects of Daily Use (London 1927) pI. LVIII. 52; Volbach 
nos. 83-85, 138; Buschhausen (supra n.1 I) B5, B13, C21; and M. H. Rutschowscaya, 
"Boites a poids d'epoque copte," La revue du Louvre 29 (1979) Iff. 

21 See Weitzmann (supra n.l1) no. 557; A. Grabar, Ampoules de terre sainte (Paris 
1958) Monza 14, Bobbio 20; Volbach no. 55. Compare also the Zachariah miniature in 



DUFFY, JOHN, A Small Box in John Moschus , Greek, Roman and Byzantine Studies, 24:1 
(1983:Spring) p.93 

98 A SMALL BOX IN JOHN MOSCHUS 

incense boxes in size and shape, but a few show scenes (such as the 
Marys approaching the altar with censers, Volbach no. 177) which are 
highly suggestive of such a use, while others show a peculiarity of 
design (a semicircular opening at the base, no. 92) which seems to 
presuppose the dispensing of incense grains.22 

Offerings. One silver pyxis of the sixth century in the St Louis Art 
Museum bears the following inscription on its lid: 1Tpoa¢opa Tt/3Ept­
v1j~ 6taKO(via~) rqJ ayicp LrEqxXvlp, "offering of the Tiberine diaco­
nate to St Stephen. "23 That the weight designation scratched into the 
base of this pyxis far exceeds its true weight suggests that the vessel 
itself was not the offering but rather the container for a more valu­
able offering, namely, a donation of precious metal to a church or 
monastery dedicated to St Stephen. Some ivory pyxides may likewise 
have served as donation boxes, which may explain why so many of 
them show one or another scene of offering or sacrifice, whether the 
Adoration of the Magi,24 the Miracle of the Loaves and Fishes (nos. 
165, 166), Habakkuk bringing food to Daniel (162, 167), the Marys 
offering incense at the altar (176, 177), the Israelites bringing offer­
ings to the Tabernacle (190), or the Sacrifice of Isaac (161-64). 

There are further possible uses for ivory pyxides. Medicine: doctor 
saints occasionally hold medicine boxes that look like pyxides~ 25 more­
over, a significant number of extant ivory pyxides show scenes of 
miraculous healing (e.g. Volbach nos. 182, 187, 194). Relics: some 
early silver reliquaries look much like ivory pyxides~ and there is no 
doubt that in both East and West pyxides were used as reliquaries in 
later centuries.26 

the Etchmiadzin Gospels: A. Durnovo, Armenian Miniatures, tf. I. J. Underwood (Lon­
don 1961) 33. 

22 See St Clair (supra n.13) 127ff; Graeven (supra n.12) 162f; and C. Schneider, 
ReallexAntChrist 1 (I950) 63ff s. v. "Acerra." The silver-gilt censer from the Kumlu~ 
Treasure (E. Akurgal, C. Mango, R. Ettinghausen, Treasures of Turkey [Geneva 1966] 
98) is decorated with an infancy cycle much like that found on contemporary ivory 
pyxides (e.g. Volbach no. 169). The censer (and presumably the incense-holder) had 
an important role in private as well as public piety: cf. P. van den Yen, La vie ancienne 
de S. Symeon Stylite Ie Jeune I (SubsHagiogr 32 [I962]) 151 *. 

23 Buschhausen (supra n.l1) C3, Weitzmann (supra n.l 1) no. 573. For 7Tpompopa as 
'donation' see P. Oxy. XVI 1899.23 n.; Preisigke s. v.; P.Ness. III 79.25 and p.227. 

24 Volbach nos. 171, 173, 173a, 199. Recall the Magi on the hem of Theodora's cloak 
in the famous imperial donation mosaic at San Vitale in Ravenna. 

25 See S. Pelekanides, Kastoria (Thessaloniki 1953) pI. 266. 
26 See Buschhausen (supra n.ll) B20; Volbach 77ff; C. Astruc, "L'inventaire dresse 

en septembre 1200 du tresor et de la bibliotheque de Patmos," Travaux et memo ires 8 
(I981) 21. 



DUFFY, JOHN, A Small Box in John Moschus , Greek, Roman and Byzantine Studies, 24:1 
(1983:Spring) p.93 

JOHN DUFFY AND GARY VIKAN 99 

The ivory pyxis and the f..WV~iKtOv may well have been one and the 
same; yet even if they were not, each enlarges our understanding of 
the other. Moreover, both are susceptible to the same definition, 'a 
small box for keeping precious objects', precisely because both were 
generic. Indeed, the design of ivory pyxides was probably determined 
much less by their function than by their medium - by the very size 
and shape of the elephant's tusk. The impression they give is one of 
series pieces-of mediocre workshop products ready to be purchased 
off the shelf.27 Their iconography may not have been intended to be 
anything more than generally meaningful, and thereby multi-func­
tional, and this may explain why generations of art historians have 
failed to make much sense of their iconography. One is inclined to 
conclude that the functions of both pyxis and f..WV,iKtOV, whether to 
hold bread, gems, incense, offerings, or something else of value, 
were probably determined, and changed, at the discretion of the 
owner. 

THE UNIVERSITY OF MARYLAND AND DUMBARTON OAKS 

January, 1983 

27 See A. Cutler, "Observations on the Production of Ivory Carvings in Late An­
tiquity and the Early Middle Ages," in Artistes. artisans et production artistique au moyen­
age (Rennes 1983) 966ff.