The Church ofSt Peter the Elder in Zadar is a complex building composed oftwo mutually related areas: the single-aisled early Christian church, which is usually called St Andrew's, and the two-aisled church called St Peter the Elder. The earliest is the single-aisled church with a hemispherical apse called St Andrew, from the early Christian period. A rectangular stnlcture was subsequently constructed to the east of this church, which we call St Peter the Elder, and in a third phase this rectangular structure was re-vaulted, at this point being outfitted with two arched aisles and two apses. The eastern half of the Church of St Peter the Elder has been included in the list of Croatian pre-Romanesque churches, and as such has become a part of various typologies ofpre-Romanesque architecture in Dalmatia, although there are no direct parallels, other than in the double apsed nature ofone part ofthe church. In fact, this in an entirely isolated example. When the very typology is thus in question, then the entire structure should be considered, the early Christian church together with the annex, and only in this manner can it be compared to similar examples. It is afact that the two-aisled space cannot in the least be considered independently ofthe western part ofthe Church ofSt Peter for a series ofreasons. First ofall, the two-aisled part was added to the eastern side of the early Christian church, allowing its apse to intel/ere in the two-aisled area. If this eastern section were to be considered a separate church, it is hardly likely that even in the most cramped spatial conditions such an infringement in the cubic shape of the two-aisled structure would be permitted. At the same time, the liturgical focus of the structure, its apse, was visually shut off, for which a velY good reason had to exist. In this sense, it is important to re-examine the reasons for the first adaptation, this being the addition of a nonetheless still undetennined first annex. On the other hand is the fact that the communication between the two areas at least at some point took place through the apse of the early Christian church, which then definitely served as the sacristy to the frontal, western section. T7Je annex in fact, in its first phase, without vaulting, had at least one door, but these were walled in when the vaulting was introduced. From that moment eJ, some other door was in function, on tbesouthern side, but also in tbe bay in front of the sanctua1Y, in the area bounded by the altar screen. In terms ofits POSition and form the eastern part ofSt Peter the Elder belongs to a p,roup ofexterior crypts. The crypt ofSt Peter the };lder thus should not be used in typologies of pre-Romanesque architecture in the Adriatic, as it is not an independent church. When crypts in this region are discussed, it should be noted that they are exceptionally rare, and that the majority were created in the Romanesque period. Ifwe exclude the C1ypt of the cathedral in Novigrad, which is otherwise directly related to Aquileia, and those in southern Dalmatia (such as St Peter in Dubrovnik), the only ones that remain in Dalmatia are those from Zadar and those created under the direct influence of Zadar. This primarily refers to two monumental crypts in the cathedrals of Zadar and Rab. The others are smaller and of various forms, but their territorial distribution is interesting. Along with St Peter the Elder, another two crypts exist in Zadar, at the churches of St Anastasia in Puntamika and St Dominique. The long-ago claim by Iveković about a crypt in the Zadar church of St Chrysogonus should also be remembered in this context. The great majority ofknown crypts are thus from Zadar or its surroundings, and most belong to the Romanesque period. It seems that the earliest are at the churches ofSt Anastasia in Puntamika and St Peter the Elder in Zadar. St Anastasia in Puntamika is an adaptation of a Roman cistem, above which the upper church was built, and despite this adaptation, it has a standardform of a twofloor structure. Inasmuch as this church can truly be connected to the legend about the conveyance ofthe relics ofSt Anastasia, then this represents evidence ofthe memorial character ofthe structure. In this sense, the crypt of St Peter the Elder cannot be related merely to the other crypts, but also with areas whose functions as a rule are derived from those that have crypts (Fig. 5). From this point ofview, the Church ofSt john in Rab is a more than illustrative example. Its deambulatory was derived from the crypt, and the memoria with a non-extrusive apse was located in the ground level of the bell-tower, just next to the sanctuary of the church on its southern side. Such a position of a belltower, noting the site ofthe sanctuary or choir, marked the holiest place in the church. Raised in the vicinity or even above the relics, they did not merely guard them in a symbolic sense, but truly in a great number of cases, small memoriae or even two-storied chapels were erected in them. At Rab, with such a spatial organization ofthe tower next to the sanctuary, the area with the apse at ground level can only be the site where relics were kept, thus an oratory-reliquary or treasury. There are several such oratories-reliquaries in Dalmatia: the memoria next to the sanctuary ofthe monastery church ofSt Peter in Osor, and the "old sacristy" ofSt Euphrasius in Poreč, the small vaulted area next to the monastery church of St Michael at Lim, and the memoria next to the cathedral at Osor. Some ofthem had another storey, which has been confirmed to date for the memoria next to the Church ofSt Peter in Osor. If the structure of St Peter the Elder in Zadar must be placed into some typology, then the only proper course would seem to be to relate it to crypts and their functional derivatives, such as orato ries-reliquaries. If it can be accepted that St Peter tbe Hlder acqUired an exterior crypt with the construction uf vaulting and two apses in the third phase of the complex, it is not entirely clear when this occurred. In this aspect, a more detailed determination ofthe period of construction and even more, the function of the second phase, when the early Christian church received a still entirely undefined annex on the eastern Side, would certainly help. Unfortunately, the minimal data cannot aid more precisely in determining a more exact chronological definition, so it is necessary to turn to stylistic and comparative material. In this sense, however, two equally possible options exist. In the construction ofarched vaulting, the use ofRoman spotia -columns, bases, capitals, and so forth -became dominant in designing the interior, with very clear signs of the "Christianization " of these elements (crosses carved on columns). And such emphasized and clearly visible use ofspotia, along with the construction technique, and the articulation of the arch friezes, can be found in Zadar only at the Church ofSt Donatus. In this sense, not even the niches on the eastern facade nor the window openings stand apart from the whole. Several fragments of interlaced sculpture found in the church, showing indications of the 9th century, should be added to this. Thus it would be very important to establish whether an altar screen existed as early as the second phase, as is suggested by part of the base of a screen pressed up against a pilaster, and then to relate the fragments of interlaced patterns but also part of the spolia to this phase. On the other hand, it should be emphasized that the articulation ofthe structure ofthe crypt ofSt Peter the Elder and the articulation of the vauIts with their strong belts is reminiscent both of the memoria in Osor and the "old sacristy" in Poreč, and finally the type ofcross-vaulting with aspherical (pendentive) finish in the oratory-reliquary ofthe Church ofSt John in Rab. Ifwe add to this that all the comparative examples of memoria come from the earlier Romanesque period, and that all crypts of the Zadar region, except perhaps that at St Anastasia in Puntamika, are Romanesque, and that the construction of exterior crypts mainly occurred in the 11th century, then this possibility is fairly well argued. For this reason, I am more inclined to leave the question of dating the third phase ofSt Peter the Elder in Zadar to future judgments. The same is true for the second phase ofthe complex -the construction ofthe annex to the Church ofSt Peter the Elder on the eastern side.