The creation of a developed chronological classification is a necessary preliminary condition which allows us to use amphora stamps of those centres where systematic stamping of ceramic containers was carried out both as reliable dating material and as a valuable source for the history of production and trade.
In the last three decades a classification scheme developed by R.B. Akhmerov was widely used in work on the magistrate stamps of Khersonesos. The author divided all the astynomes known to him into four chronological groups having established absolute dates for each of them. This classification, in spite of a number of serious critical comments made by A.A. Neichardt, continued to be considered quite reliable until most recent times.
In the meantime relatively recently there has appeared material which does not fit the scheme proposed by R.B. Akhmerov. This material was partially used by B.Iu. Mikhlin in the course of work on re-dating chronological groups, sub-groups, and individual Khersonesan stamps. At the same time, the proposed corrections are so serious that they do not really fit with Mikhlin's conclusion that so far there are insufficient grounds for doubting the correctness of the grouping of Khersonesan magistrates as proposed by R.B. Akhmerov. In our opinion these grounds not only exist but allow us to raise the question of the necessity of a radical revision of this classification.
The point is that erroneous datings are far from being its main short-coming. We should recognise that the artificial character of the individually constructed chronological groups themselves is a much more substantial fault, as a result of which, as A.A. Neichardt correctly observed, "These [p88] groupings and groups are so close that R.B. Akhmerov is not able to define clearly the difference between them."
There is no doubt that in the course of his work, R.B. Akhmerov had to face considerable difficulties caused not only by insufficient [previous] research but also by specific features of the Khersonesan magistrate stamps. The almost complete absence of devices in them excluded the possibility of applying numismatic methods in grouping and dating the impressions. The rare usage of a second name made it difficult to apply a synchronistic method for these purposes. As the author himself notes, in the creation of his classification he relied on "... following material: 1) Khersonesan coins with the names of magistrates, 2) lapidary monuments with names which coincide with the names on ceramic stamps, 3) archaeological complexes which make it possible to establish the date of the stamps, 4) palaeographic and orthographic features of the stamps." The shapes of the dies, patronymics, and monograms on individual stamps were also taken into account. It would seem that the use of such a wide range of material should lead to the creation of a valuable classification. However, this did not happen, and could not have happened.
The initial methodological principle used by R.B. Akhmerov turned out to be wrong and inadmissible --- striving to determine temporal frames of the activity of each of the astynomes in order to go on to unite magistrates close in time into chronological groups. In this way the author started dating individual stamps before he created their typological classification. In connection with this he did not develop and clearly isolate the features which would have allowed him to group stamps of individual magistrates objectively.
Moreover, R.B. Akhmerov did not always clearly visualize, and frequently overestimated, the possibilities of the methods which he applied in the course of his work on creating a classification scheme. Thus it remains unclear what he understood under "generally accepted methods of palaeography." This rubric needed clarification particularly because with the rather conservative nature of the letter-forms of ceramic inscriptions their palaeographic analysis usually plays an auxiliary role and is used only for relative confirmation of the synchronicity of different groups of stamps singled out by other methods. It remains unclear what the observations of "style of inscriptions in their historical development" provided for the author. The orthographical method for the purposes of classification usually also finds only a limited application. R.B. Akhmerov also did not have at his disposal sufficiently reliable archaeological complexes containing Khersonesan stamps.
At the same time in forming chronological groups R.B. Akhmerov devoted a disproportionately large place to the method of dating individual astynomes by coincidence of their names with names contained on coins and lapidary monuments. The author clearly ignores the fact that positive results can be obtained only in those rare cases where not only the names but also the patronymics coincide in stamps, coins, and inscriptions. When only identical names are present, their identification always remains a guess.
Thus the set of methodological means used by R.B. Akhmerov in creating a chronological classification of Khersonesan magistrate stamps makes it possible to determine absolute dates of the activity of individual astynomes and to make the relative chronology of some groups of stamps more precise. However these means could not at all and should not form a base for the actual grouping of the dies. The latter should be based on a developed typology of the stamps themselves. R.B. Akhmerov for all practical purposes did not do work in this direction. This is precisely why the classification proposed by him turned out to be imperfect and did not pass the test of time.
A successful attempt at a typological analysis of the magistrate dies of Khersonesos was carried out for the first time only by B.Iu. Mikhlin. The summary tables of his article provide statistical material which characterizes certain tendencies in the changes in the shapes of the stamps, the contents of the readings, and the way the magistracy is written on them. At the same time, B.Iu. Mikhlin kept unchanged the basic principles of the old grouping. As a result the tables given by him once more confirm the fact that dies of different types were included in the composition of each of the chronological groups without adequate grounds. In connection with this the development of precise typological classification of Khersonesan magistrate stamps remains a task of the first priority, and it should begin with the clarification and correct grouping of the unaccidental features which characterized them.
There is an opinion that the stamps of Khersonesos are quite uniform. Indeed, in comparison with the impressions on the amphoras of Thasos, Sinope, Heraclea, and even Rhodes, they are more homogeneous. However, even among them we can succeed in isolating several types, characterized by a statistically stable combination of a number of features. Out of them the following are the most important for classification: shape and dimensions of impressions, composition and structure of the readings on them, palaeographic and orthographic peculiarities of the inscriptions. We will try to single out the more and less essential features from this set, i.e., reveal their hierarchy.
At the time of writing this article we have already collated more than 2.5 thousand (about 70% of all those known) Khersonesan magistrate stamps. It turned out that these impressions were made with 370 dies.
Shape. According to this feature, stamps are divided into grooved, weakly grooved, flat, and figured. Stamps which are grooved --- deepened and semi-circular in cross-section --- are considered most typical for Khersonesos. Indeed this shape has no analogies among dies of other centres. Moreover, two-thirds of the dies presently registered produced impressions of grooved shape. Experiments have shown that these dies were most likely engraved on the lateral surfaces of amphora handles. The unique shape of the dies resulted in the fact that full impressions came out extremely rarely. Frequently in two line stamps the tops of the letters of the upper line or the bottoms of the letters of the lower line are smudged. In three-line impressions, the upper or lower lines are often fully smudged.
Flat rectangular stamps (at present more than a hundred dies which were used to produce these stamps are known) usually are the best preserved. Their characteristic feature is the presence of a "frame" --- clear borders on all four sides of the stamp. So far it is not clear what material was used to produce the dies for flat stamps. It is quite possible that sherds from ceramic artifacts with a flat surface were used for this purpose. Fragments of amphora handles with a filed lateral edge could have been used as well. In that case the existence of individual examples of stamps (they are impressed with seven dies) becomes understandable, as they were a transitional form between grooved impressions and flat ones. They have a clearly observable "frame," as in the flat stamps, as well as slight concavity, which brings them closer to grooved ones.
Two dies with the name of the agoranome Apollonidas, which produced figured flat impressions, should be recognised as the original creative work of one particular die-cutter.
Calculations show that out of 121 names of magistrates known at present, 38% are found only on grooved stamps; 45% only on flat ones; 11% both on flat and grooved ones; 5% on flat, grooved, and weakly grooved; 1% on figured impressions.
The dimensions of the stamps are quite varied: the length of grooved impressions fluctuates within the limits of 4.6--8.8 cm, flat ones within 2.9--9.9 cm; the width of grooved stamps measures 1.2--2.0 cm, flat ones 1.2--1.6 cm. The fact that the dimensions of the stamps do not function as independent features causes no doubts, because their values are firmly correlated with such palaeographical aspects of each inscription as the height and width of its letters, as well as the variant of the stamp reading, which determines the number of lines and the number of letters in each line.
Reading. Other features characterizing magistrate stamps are connected with unique traits of the inscriptions which they bear. Inscriptions are usually set in two--three lines. Four- and five-line dies are found considerably more rarely.
The majority of the stamps have a reading which consists of two main elements, united by the die-cutter in one die: name (or name and patronymic) and the name of the magistracy. There are two variants of arranging these elements: a) the name of a magistrate is placed first in the inscription, b) the first place is given to the title. Each variant, in its turn, has a number of different variations, determined in relation to the features in the line-by-line distribution of the basic elements of the reading.
Stamps where the name of an official is not accompanied by the
title are found relatively rarely (only 7% of magistrates). More
frequently (12% of magistrates) stamps exhibit the third component of
a reading --- a monogram or an abbreviation of the second name, usually
understood to be the name of the potter. Stamps of the astynome
Athaneus, which contain, apart from the name, either the adjective
The readings of the magistrate stamps known to us vary within the limits of the following combinations: 1) name, title; 2) name, title, monogram, or abbreviation; 3) name, adjective; 4) name with patronymic; 5) name with patronymic, abbreviation; 6) name with patronymic, title; 7) name with patronymic, title, abbreviation; 8) name with patronymic, title, device; [p.91] 9) title, name with patronymic; 10) title, name with patronymic, monogram.
Orthography. Linguistic features, peculiar to the inscriptions of Khersonesan stamps, made it possible to define their dialect as one of the Megarian offshoots of the Dorian dialect. The following regular characteristics in the writing of both the magistrate names themselves and the titles are of particular interest for the typological analysis of the impressions.
Orthographic mistakes are exceptionally rare in the stamps. We know only eight dies, the inscriptions of which have either an accidental reversal of letters or one letter accidentally missing.
Palaeography. Analysis of the palaeographic features of Khersonesan ceramic inscriptions shows the closeness of their letter-forms to the letter-forms of lapidary monuments. This explains the strong conservatism, severity, and certain "elegance" of the alphabet. The influence of the hand-written writing is practically limited to the frequent use of the lunate forms of epsilon and sigma and occasional use of cursive omega.
In his time, B.N. Grakov defined two types of letters depicted in Khersonesan stamps. The first consisted of wide, squarish, even letter-forms, characteristic first of all for grooved impressions; the second, small and compressed, is usual for flat stamps. R.B. Akhmerov, in his turn, attempted to establish characteristic traits of palaeography for each of the four chronological groups singled out by him. However, this division into fractions caused justifiable objections from A.A. Neichardt who at the same time proposed to add to the groups defined by B.N. Grakov yet another group, which formed a transitional link between the first and the second.
Instances where inscriptions run from right to left are also found on Khersonesan stamps. Usually such deviations from the norm are explained by die-cutters' mistakes, who, it has been claimed, were accidentally making positive inscriptions on dies instead of negative ones. There are, however, grounds to consider that there was a conscious intention to create such dies. It is characteristic that all five dies with the name of the astynome Alexandros produced retrograde impressions. The dies are similar in the dimensions and palaeographic features of the inscriptions, which makes it possible to regard them as the production of one craftsman. Six dies of the seven with the name of the astynome Apollatheos contain positive inscriptions made apparently also by a single die-cutter.
Combinations of the values of the features described above produced a little over 70 different variations of Khersonesan magistrate dies.
Forty-eight of them have been selected for a typological analysis. Other variations differ from those selected only by the shape of the stamps or palaeographic features of the inscriptions, i.e., by features with little informative value. It turned out that the composition and distribution of the basic elements of a reading constitute the typological features which determine the most important changes within the entire category of magistrate stamps. The number of lines, the nature of the line-by-line distribution of the individual elements of a reading, and the orthographic features of the inscriptions form variant features which reflect secondary changes within a category.
Type I (see Fig. 1) unites stamps of 15 variations, grouped in five variants. A common feature, characteristic for all dies of the type, is the composition of the inscription according to the following scheme: name without patronymic occupies the first place, and the title the second. Such a scheme is characteristic for stamps of 39 astynomes and one agoranome (34% of all Khersonesan magistrates) made by 210 dies (approximately 56% of all known at the present time).
Variant 1 is represented by two-line stamps in the inscriptions of which one finds a very much abbreviated name of the magistracy, and in variation 1a the name of the astynome is also abbreviated.
Variant 2 embraces two-line stamps in the readings of which a full form (variation 2a has it slightly abbreviated) of the genitive of the noun is used in writing the title. The variant includes more than 80% of the dies of type I.
Variant 3 presents the next step in the development of the reading connected with the use of the title in the form of the genitive of the participle in two-line stamps.
Variants 4 and 5 are represented by stamps, the readings of which are analogous in their composition to the readings of the two preceding variants. However, the name of the magistracy and, in variant 5, the names of the magistrates, are divided between the two lines of the inscriptions. Stamps of the 4th variant have two lines, of the 5th, 3-lined grooved, and 4-lined figured.
Most stamps contain monograms and abbreviations in the composition of the readings.
The typological unity of all the noted variations of stamps raises no doubts. It is indicative that 9 astynomes (22% of the total number) are represented by multi-variant stamps. Approximately 84 dies produced grooved impressions, and flat stamps reign without any competition only in variant 1.
The stamp inscriptions are done in calm and beautiful letter forms. The scattered formation of the letters with a wide style of writing is typical. A height of 0.5--0.7 cm is most widely used. The width of most letters is slightly larger than, or equal to, their height. Omega, theta, and omicron as a rule are slightly smaller than other letters. The lunate form of sigma is characteristic only for the stamps of variants 2 and 3, although it is present only in one die out of every five. The cursive omega is found only in one die of the astynome Pasion.
Only 3 (2b, b', and 3b') variations may be considered as predominant and typical. Other rare variations comprise the periphery of the type.
Type II (see Fig. 2) unites stamps of 17 variations grouped into 6 variants. The main type-forming feature is the composition of the inscriptions in the stamps by the following scheme: name with patronymic first, title second. The type is represented by stamps containing the names of 53 astynomes and 1 agoranome (45% of all known Khersonesan magistrates) and made with 108 dies (approximately 29% of their total number).
Variant 1 is known from the stamps of only 3 magistrates in the inscriptions of which the name of the official is in the nominative and the article in front of the patronymic is missing. The title in 2 (1a, b) variations is given in abbreviated form, which brings these stamps close to the impressions of variants 1 and 2a of type I. In the variations 1b', g, the titles are done in the full form of the nominative of the participle.
Variants 2 and 3 reflect the culminating stages of the development of the readings of two-line stamps. The name of the astynome, always in the genitive, is usually combined with the patronymic by an article, which is omitted only in stamps of 2 [p93] astynomes (variation 3a). In the writing of the name of the magistracy, the genitive of the participle is predominate, and the genitive case of the noun is characteristic only for the stamps of variation of 2a. The difference between variants 2 and 3 lies in the different line-by-line distribution of the magistrates' patronymics; in the 2nd variant the patronymic occupies the end of the upper and the beginning of the lower line of the inscription, and in the long narrow stamps of the 3rd variant it fits fully on the upper line.
Variants 4 and 5 reflect further stages of the development of the readings of 3-line stamps. All dies are characterized by a clear division of the basic elements of the reading among the individual lines of the inscription. If, in the stamps of variant 4, the article before the patronymic is omitted (which unites them with the impressions of variant 1 of the present type), and the name of the magistracy is in the form of the genitive of the noun, then, in the stamps of the 5th variant, the article is compulsory, and the title is in the genitive of the participle. The presence of abbreviations in the composition of the stamp inscriptions of variants 4b and 5b' attracts attention and brings them close to the dies of the 4th variant of type I. Finally, among the stamps of variant 5 there are impressions in the area of which devices are present (variation 5g).
Variant 6 is the culminating stage of the development of the readings of 3-line stamps of type II. A common feature for all the impressions of this variant is the appearance of the initial syllables of the title at the end of the 2nd line from the bottom of the inscription. It is indicative that the analogous line-by-line distribution of the title first appeared already in stamps of variants 4--5 of type I. It was also used in 2-line dies of variant 2 of the present type.
Thus the stamps of type II are more varied than the impressions of the preceding type. This is connected with the parallel existence of both 2- and 3-line stamps in it. At the same time all these stamps are genetically connected with the variations of type I. Therefore one can observe similar characteristic features, noted already in the preceding type, in the development of the readings. This concerns the line-by-line distribution of the individual elements of the readings and the gradual replacement of the genitive case of the noun by the genitive of the participle in the name of the magistracy. A brief period of using the nominative in the writing of the name and title is characteristic of the dies which clearly constitute a transitional link between stamps of type I and II. The practice of omitting the article between the name and patronymic belongs to the same period. The third component of the reading --- monogram or abbreviation of the 2nd name --- continues to appear sporadically in stamps of type II. As for the shape of the stamps, one can observe a process of gradual replacement of grooved impressions by flat ones. Of the dies of type II known to us only 44% produced grooved impressions and already 52% produced flat ones. Among the stamps of the late variants 5 and 6, flat impressions already comprise more than 70%. The remaining 4% of the dies produced weakly-grooved stamps not found earlier.
In the palaeography of stamps of type II one can observe the decrease in the dimensions of the letters, along with the preservation in general of the shape characteristic for impressions of the preceding type. The most wide-spread height of the letters is 0.4--0.5 cm. They become narrower. Omega, theta, and omicron are now considerably smaller in dimension than the other letters. The lunate form of sigma becomes as popular as the broken one (it is found in 46% of the dies). Cursive forms of epsilon and omega are used more widely. In some stamps the letter-forms are uneven and hastily written.
Variations 2a, 3b, 5b, and 6b', are predominant. The rest represent rare shapes comprising the periphery of the type.
Type III (see Fig. 3) unites stamps of 8 variations, grouped in 3 variants. A common feature for all dies of the type is the unique scheme of the arrangement of the inscription, where the name of the magistracy is placed first, and the name with patronymic second. Such a distribution of the basic elements of the reading is characteristic for the stamps of 33 astynomes (27% of all known) made with 46 dies (approximately 12% of the total number).
Variant 1 is genetically connected with variants 3 and 5 of the preceding type. The difference lies in the transfer of the name of the magistracy from the last to the first line of the inscription. The rare variations 1a, b' complete, the former the evolution of 2-line magistrate stamps, and the latter the practice of including monograms in the composition of the dies.
Variants 2 and 3 represent the final stages of the development of the readings of magistrate stamps. Typologically their dies came from the impressions of variation 1b of the same type. The decrease in the length of the dies caused a reduction in the number of letters in the lines of the inscriptions which resulted in the necessity of transferring the ending of the magistracy name to the 2nd line (variant 2) and later the placing of the title fully in the first 2 lines of the inscriptions (variant 3).
In comparison with the preceding stamps, the stamps of type III have a smaller number of variants and variations. 1b, 2b, 3b are predominate and comprise the nucleus of the type. Other variations are rare. The homogeneity of all stamps of the type raises no doubts. In this type the process of replacing grooved impressions with flat ones culminates already with the astynomes of variant 1. The inscriptions of type III stamps are distinguished by small, clear, and compressed letter-forms. The average height of the letters is 0.30--0.35 cm, and their width is either equal or slightly larger than their height. Omicron is usually smaller than other letters, and frequently, instead of a circle, is transformed into a dot of 0.15 cm in diameter. The general shape of the letters characteristic of the preceding type is preserved, but a considerably greater variety in their writing is observed. Thus the dot of the theta is frequently replaced by a bar. Cursive omega is as popular [p96] as its earlier shape. Various ways of writing sigma are found: it can be widely opened, with parallel upper and lower hastas, and lunate. Frequently the letters are decorated with apices at the ends.
Type IV (see Fig. 4) unites stamps of 6 variations grouped into 3 variants. The absence of the title in the composition of the readings is a type-forming feature. The type is represented by impressions which contain only 7 names (6 of which are accompanied by patronymics) and made with 13 dies (3% of all known).
One has to keep in mind that far from all researchers have attributed these dies to magistrates. Already in 1949, V.V. Borisova put forward the supposition that they belonged to the owners of workshops. Later, when publishing a list of Khersonesan astynomes, she defined such stamps as "magistrateless". However there are weighty grounds to suppose that the name of the magistracy is simply omitted in these impressions. The following observations confirm this deduction. In the composition of the readings of the stamps of variation 2b there are abbreviations which are usually defined as abbreviated names of craftsmen or owners of ceramic workshops. Analogous abbreviations are also found in small supplementary stamps which accompany the main impressions of variation 2a. Thus there are grounds to suppose that in such cases the stamps containing names with patronymics are magistrate stamps. One cannot consider as accidental the fact that, of the 7 names represented in stamps of type IV, 4 are known to us also by dies of other types where they are accompanied by the title (see Fig. 4).
Variant 1 unites 2-line grooved stamps where in the first line
there is the word
Theogenes, son of Apollonidas, and of Herokas, son of Herotimos, in the presence of abbreviations and in the absence of the article in front of the patronymic are also close to the early impressions of type II variant 4. In connection with this there is a basis for supposing that Herokas son of Herotimos and the astynome Herokas, known on stamps of variation 2b' of type I are the same person.
Variant 3 unites stamps which preserve, as with the impressions of the previous variant, names and patronymics alone. However in them the article precedes the patronymic. Variation 3a is represented by 2-line grooved stamps of the magistrate Polystratos son of Xenon. Variation 3b is known so far only in one 3-line impression, [p98] containing the name of Neupolis son of Menestratos, while variation 3b' is known in a 3-line stamp with the name of Pythodotos son of Damokles. The last two magistrates are also known on astynome dies of the 2nd variant of type II and the 3rd variant of type III (see Fig. 4) respectively.
All categories of type IV are quite rare. In fact they could be considered as atypical variations of the three preceding types, within the limits of which the evolution of the basic features characteristic of Khersonesan magistrate stamps took place.
Firstly we can chart a distinct tendency in the development of the shapes of the impressions themselves --- there is a transition from grooved (type I) to flat (type II) stamps. Secondly, in the process of type formation the composition and the distribution of the basic elements of the reading change. The initial stage of stamping is connected with 2-line dies where the inscription consists of the name standing in the first place and the title (type I); then along with 2-line there appear 3-line inscriptions where the distribution of the basic components of the reading remains the same but the name of the magistrate is accompanied by a patronymic (type II); finally, the name of the magistracy moves to the first place in the inscriptions, which may be both 3- and 4-line (type III). At a certain period in the last stages of the existence of the impressions of type I and the initial stages of type II, the practice of including monograms and abbreviations into the composition of the inscriptions became widespread.
Thirdly, two stable forms of writing the names of the magistracy have been recorded: in the stamps of type I the full form of the genitive of the noun dominates; in the impressions of type II one can observe its gradual replacement by the genitive of the participle; the latter writing is fully dominate in the dies of type III. The abbreviating of the magistrate's name and title and the expression of the latter through a participle in the nominative are apparently characteristic for the transitional period between type I and II. The practice of omitting the article before the patronymic belongs to the initial stage of type II.
The tendency to decreasing the dimensions of the letters in the inscriptions is clearly traced throughout the entire period of stamping. If the palaeography of the stamps of type I finds its closest analogies in the lapidary monuments of Khersonesos of end 4c--1/2 3c BC, then the dies of type III already have the features characteristic of inscriptions of end 3c--2c BC.
It is interesting that the observed tendencies in the development the of readings of the magistrate stamps of Khersonesos have undoubted similarities of feature with the Sinopean epigraphic tradition. In the astynome stamps of Sinope the composition of the reading initially also had only the name of the magistrate (chronological groups I--III) which usually occupied the first line of the inscription, and the title in the genitive case of the noun is in the second place. Later one can observe as in Khersonesan stamps the appearance of the patronymic with the name of the astynome and a gradual transition to writing the magistracy in the genitive case of the participle (groups IV--V). Here as well the evolution culminates in placing the name of the magistracy in the first place and its division between the two first lines of the reading (groups V--VI).
Thus the supposition that the ceramic production of Sinope influenced not only the morphological peculiarities of Khersonesan amphoras, but also the typology of the impressions placed on their handles and necks, is highly probable.
The correctness of the proposed scheme of the typological development of Khersonesan magistrate stamps is well illustrated by the table given in Fig. 5. It groups samples of dies belonging to the members of three families, who systematically occupied the position of astynome for the duration of four to six generations, i.e., 100--150 years consecutively. In all the family groups one can observe a successive transition from using stamps of type I via using impressions of type II to the appearance of stamps of type III.
These observations strengthen the conclusion that the impressions of types I--III represent a continuous typological series which at the same time is a chronological series. Therefore each of these types may form the nucleus of one of three chronological groups.
Until very recently the generally accepted chronological borders of astynome stamping in Khersonesos were end 4c and beg 1c BC, i.e., a period of two and a half centuries duration was focussed on. However by the present time we know of only 121 magistrates, and this fact right from the start forces us to doubt the accuracy of the proposed borders. Therefore the attempt of B.Iu Mikhlin to make the chronological frames of the production of Khersonesan stamped containers more precise should be acknowledged as legitimate. But if his proposal of the cessation of the practice of stamping in the middle of the 2c BC deserves attention and may serve as a starting-point for further analysis, still his moving of the lower border earlier to 330 BC in our opinion has no justification. This date is refuted by the absence of stamps of Khersonesan astynomes in the composition of the ceramic complex obtained recently in the course of studying a cultural layer from under the ruins of an ancient theatre in Khersonesos. The upper border of the complex is firmly determined as 320 BC. In connection with this, end 4c remains the most probable time for the appearance of the practice of stamping Khersonesan containers in Khersonesos, while 320--315 may be provisionally accepted as the lower chronological border of the classification of Khersonesan stamps.
Chronological Group 1 (end 4c--1/4 3c BC). The typological analysis of magistrate stamps conducted above shows that there are grounds for connecting the first period of stamping ceramic containers in Khersonesos with the appearance and existence of impressions of type I.
This deduction ties in with the results of studying collections of Khersonesan stamps discovered in the last decades in excavations of settlements where life ceased not later than mid 3c BC (see Table 1). This dating makes it possible to propose that in the composition
Table 1 Complexes of Khersonesan stamps end 4c--1/2 3c BC -------------------------------------------------------------------------- | Total | Including | number |----------------------------------------- Centre |  | Type | I | II | IV | |---------|------|------------------|----- | | Variant | 1--5 | 1 | 4 | 5 | 2 ---------------------|----------|---------|------|------------------|----- Roksolanskoe ancient | | | | | | | town (Nikonii) | 15/25 | | 9/17 | 3/4 | 2/3 | 1/1 | -- | | | | | | | Panskoe-I (estate nr | 13/73 | | 10/70| -- | 1/1 | 1/1 | 1/1 6) | | | | | | | | | | | | | | Elizavetovskoe set- | 17/41 | | 15/38| 1/2 | 1/1 | -- | -- tlement | | | | | | | --------------------------------------------------------------------------
of the complexes being analysed there should be stamps of the early chronological groups. Indeed the impressions discovered here are of exceptional typological homogeneity: from 70% (at the Roksolanskoe ancient town) to 95% (at the Panskoe-I settlement) of the stamps belong to type I; isolated examples belong to the early variants of type II and variant 2 of type IV, which is close to them in time.
Thus the basic bulk of the production in ceramic containers arrived at these settlements from Khersonesos under the astynomes of chronological group 1 (type I of stamps) and the cessation of export coincides in time with the transition in Khersonesan ceramic stamping to the dies of chronological group 2 with which the impressions of types II and IV have already been connected. Taking this into consideration the upper chronological border of group 1 could be made more precise by determining the time when life ceased at these settlements. For Roksolanskoe ancient town this date is established in the relatively wide limits of 2/4 3c and for Elizavetovskoe settlement no later than the middle of the century. The closed ceramic complex from estate nr 6 of Panskoe-I settlement gives more precise borders: end 4c--1/3 3c BC. The latest ceramic stamp discovered here belongs to the astynome Prytanis son of Ariston (variant 5 of type II) whose activity, judging by the discovery of an analogous impression in the Zelenskii tumulus, also does not go beyond the limits of the 70s of 3c BC. And since this astynome is not the earliest among the magistrates who produced stamps according to type II, the final date of chronological group 1 should be moved at least to the middle of the 70s of 3c BC.
The list of magistrates known by stamps of type I as we have seen has 39 names of astynomes and 1 agoranome. However we should exclude three officials from this number and place them into the next chronological group: astynomes Herokrates and Herokas, known by dies not only of type I but also of types II and IV respectively, and agoranome Apollonidas who is, apparently, close in time to agoranome Apollonios son of Pasiadas.
One cannot exclude the possibility that there are homonyms among the 37 astynomes remaining in the list. However, an attempt to isolate them is difficult owing to the relative typological unity of group 1. Even when one comes across stamps of different variants containing the same name there is no complete assurance that these are homonyms.
When we published a more precise list of Khersonesan magistrates, we put forward a supposition that there existed 3 pairs of astynomes which were close in time and bore the names of Agasikles, Hereas, and Herodotos. One must say that there were grounds for such a conclusion. On the one hand, all [p102]
3 names are known in stamps made according to the first variant using a strongly abbreviated name of the magistracy. If one was to define this feature not as an accidental abbreviation but as a reflection of the old orthography of the genitive case of the second declension, then it is possible to suppose that these dies belong to an early stage of stamping. On the other hand the same names are represented in the impressions, the inscription on which is carried out according to variant 2 using sigma lunate and, particularly important, according to variant 4, which contains features characteristic, as we have seen, of late dies of type I. We tried to resolve this contradiction by putting forward the supposition that there were magistrates who produced stamps according to type I who were homonyms [? punctuation emended. tr]. However at present we have to abandon this hypothesis.
B.Iu. Mikhlin published a tile which contained 2 astynome stamps: a flat one with the name of Herodotos and a grooved 3-line one with the name of Agathon son of Gnathon. The very fact of placing stamps of different magistrates on one product is undoubtedly unique, and may be explained only by an accident connected with the presence of two different name dies in the hands of the potters during the transition of astynomes. For us this lucky find is important for two reasons: first of all, it confirms the deduction on the gradual replacement of type I stamps by impressions of type II in the ceramic epigraphy of Khersonesos, because the existence of the period of parallel co-existence of different types of stamps has been recorded; secondly, the activity of the astynome Herodotos should be determined by the end of chronological group 1. Meanwhile, according to the shape, the presence of the abbreviated name of the magistracy, and palaeographical peculiarities, this stamp of Herodotos is similar to other dies of variant 1 which have the same name. One can observe an even greater similarity, including that of the dimensions, to the stamps of the astynome Xenon which also belong to variant 1. In connection with this one can suppose that all the impressions of variant 1 are chronologically connected with the end of group 1, and thus coincide in time with the stamps of variant 4. Consequently, there are no sufficiently weighty grounds for considering the astynomes, who have the same name, and stamps of both the first and the fourth variants of type I, as homonyms.
At the same time we can suppose that the homonyms were 2 astynomes who were close in time and bore the name of Herakleas. This is evidenced by the existence of 18 dies containing this name, while the average number per astynome of the group equals only 5 examples.
Taking all of the above into account, 38 astynomes were included in chronological group 1, and the period of its existence is determined as 40--45 years. The relative typological unity of the group makes it difficult to divide it into smaller stages limited in time. However, taking into account the logic of the development of the stamp shapes and the production of the inscriptions on them, it is possible, even if in a preliminary way, to divide all the magistrates into 3 consecutive sub-groups. The representatives of the later sub-groups are easier to determine because in their stamps there are features which had just appeared in the process of evolution of the dies of type I and which became dominate later: widespread usage of flat impression together with grooved ones, the appearance of the name of the magistracy on them in the genitive of the participle, the presence of monograms and abbreviations of second names in the area of the stamps and the usage of cursive shapes of letters in the inscriptions.
Sub-group A includes 14 astynomes, whose stamps lack the features mentioned above characteristic of later impressions of the group. All the stamps are of the grooved shape, and done with the dies of variation 2b' of type I.
Sub-group B includes 15 astynomes, known by stamps of 4 variations of variants 2 and 3 of type I. As in the preceding sub-group, the dies of variation 2b' occupy the dominate place here. However, in the inscriptions of the sub-group, the lunate sigma has begun to be widely used alongside the broken one. The practice of writing the magistracy in the form of the genitive of the participle arises. Flat stamps become popular alongside the grooved ones.
Sub-group B' includes 9 astynomes represented by the stamps of now 7 variations of type I and 1 variation of type IV. Such a typological kaleidoscope is characteristic for the periods transitional between chronological groups. The astynomes Agasikles, Hereas, and Herodotos comprise the nucleus of the group. Flat impressions of the astynomes Xenon and Philippos are close in their dimensions to some stamps of the magistrate Herodotos. The magistrates Damoteles, Dioskourides, and Polystratos are united by the presence of monograms on their stamps (variations 2b and 3b), which brings these impressions close to stamps of variant 1 of type I. Finally the sub-group includes the astynome Athanaios in whose stamps we first come across the practice of omitting the name of the magistracy (variant 1 of type IV). Already, 34% of the dies in the sub-group produce flat impressions. The practice of abbreviating titles became relatively widespread. Almost half of the dies have monograms and abbreviations in the composition of the readings.
In conformity with the number of astynomes in each of the sub-groups, and taking into consideration the chronological framework of the entire group we can tentatively determine the following temporal borders of each of the stages: sub-group A end 4c, sub-group B first two decades of 3c, sub-group B' end 80s--mid 70s 3c BC.
Chronological Group 2 (275--215 BC). The nucleus of the group is comprised by 42 magistrates, known only according to stamps of type II. Apart from them, the composition of the group includes 3 more officials represented by stamps of type I and 4 magistrates which have impressions of variants 2 and 3 of type IV, close, as we have seen, to the early variants of type II. Finally the astynome Bollion son of Nikias, one of whose dies was done according to type III, was placed into this group. Thus the total list of the group consists of 50 magistrates (see Table II, III).
The lower border of chronological group 2 has been determined by us earlier [p104] as the middle of the 70s of 3c. Due to the absence of reliable archaeological complexes, which contain stamps of the end of this group, it is difficult to make its upper border more precise. However, a number of indirect data help to find a certain point of departure. Thus among astynomes of the transitional period from the 2nd to 3rd chronological groups, i.e., among magistrates represented by stamps both of type II and of type III, there is a certain Hymnos son of Skythos. This name moreover was placed on Khersonesan coins of 4/4 3c, in a deed on a sale of land, and, particularly important, among Delphian proxenies of 195/194 BC. Researchers unanimously consider that in all these cases we are dealing with the same person. Obviously B.Iu. Mikhlin is completely right to place the occupancy of the astynome position, one of the lesser magistracies in Khersonesos, by Hymnos at the beginning of the last decade of 3c, i.e., 10--15 years earlier than the time when he became a Delphian proxenos. Thus there are grounds for determining the border of the transition from the 2nd to 3rd chronological groups as a time near 215 BC.
Taking into account the regular laws observed in the development of the stamps of type II, we have the possibility of dividing the astynomes of chronological group 2 into 3 stages.
Sub-group A includes 14 magistrates known by stamps of 4 variations of type I, 10 of type II, and 3 of type IV. In its typological aspect the composition of the sub-group is extremely diverse. This is explained by the fact that the impressions of the sub-group belong to a transitional period in the course of which the emergence of the new type, type II of magistrate stamps, took place. The process of replacing flat impression by grooved ones continues. In connection with the appearance of patronymics with the magistrates' names, the 3-line inscriptions replace 2-line. They are already characteristic for 60% of the dies in the sub-group. In the readings of 19% of the dies the name of the magistracy is omitted, and the remainder exhibit a lack of unity in the way they are written. Together with the previously known forms of the genitive of the noun and participle in dies which contain the magistrate's name in the nominative, sometimes an abbreviated title appears or it is written in the nominative of the participle.
In spite of the typological kaleidoscope there are certain features of similarity among individual variants of the sub-group's stamps according to the composition of the readings and the line-by-line arrangement of the inscriptions on them. Moreover in the inscriptions of 11 magistrates, such a characteristic feature, typical for this sub-group, as the omission of the article joining the name with the patronymic has been recorded. The chronological unity of the magistrates of the sub-group is confirmed by joint finds of stamps of the most wide-spread variants 1 and 4 of type II, and of variant 2 of type IV, in the previously-mentioned complexes (see Table I).
The stamps of variant 5 of type I with the name of the astynome
Prytanis son of Ariston stand alone within the sub-group. Other
magistrates, who have one-variant dies, are ascribed by us to the next
sub-group. An exception was made for Prytanis. First of all, only this
astynome is known by stamps made according to variation 5b' and
containing in the composition of the reading, an abbreviation of the
2nd name in the form of
Sub-group B includes 17 astynomes known by stamps of 6 variations of type II and 1 of type IV. Typologically the sub-group is more homogeneous than the preceding one. It is no accident that its nucleus consists of the stamps of only two (2a and 5a) variations of type II. Already half of the dies of this sub-group produced flat impressions. The writing of the titles in the genitive of the participle takes over completely. The number of stamps where cursive shapes of letters are used in the inscription is again increasing.
Sub-group B' contains 19 astynomes known by the stamps of 8 variations of type II and 1 of type III. The nucleus of the sub-group consists of the stamps of two (3b and 6b') variations of type II. The process of supplanting grooved stamps by flat continues. At the same time weakly grooved forms appear. The writing of the name of the magistracy in the genitive of the participle takes over completely. In the readings of 2/3 of the dies, lunate sigmas are found.
The astynome Bollion son of Nikias, one of whose dies was done according to variation 2b of type III stands alone in the sub-group. At the same time 2 dies of this magistrate produced impressions according to type II. If one of them is attributed to variation 6b', which is the norm for sub-group B', the second one is attributed to an extremely rare variation, 5g, known by stamps of another 3 astynomes, which belong to the preceding sub-group B. All the dies of variation 5g are apparently done by the same die-cutter and are close in time. Therefore there are grounds for leaving astynome Bollion in the composition of chronological group 2.
Provisional borders of each of the three sub-groups within the limits set for chronological group 2 are determined by taking into consideration the number of magistrates assigned to each one of them. For sub-group A this would be the end 70s--60s, for sub-group B 50s--beg 30s, sub-group B' end 30s--beg 10s 3c BC.
Chronological Group 3 (215--60s 2c BC). The remaining 33 astynomes known by stamps of type III are included in this group. At present its upper border is hard to establish because there are no at all complete and well-dated ceramic complexes containing later Khersonesan stamps.
At first glance taking into consideration the fact that little more than 30 astynomes of this group are known, the cessation of the practice of stamping ceramic containers in Khersonesos could have been attributed to end 1/4 2c BC. However, as we see it, such a dating is too low by at least 10 or 15 years. It is necessary to take into account the possibility of short-term intermissions in the period when amphora stamping in Khersonesos was dying out. It is indicative that most of the astynomes in the groups are represented by only a few, and frequently only by 1--2, impressions. In connection with this, the probability that a number (apparently a considerable number) of the magistrates of the last period of stamping still remain unknown to us is quite high.
From the time of the publication of the list of Khersonesan astynomes compiled by R.B. Akhmerov, 11 new magistrates have been discovered. Of them, 5 belong to chronological group 2 and 6 to group 3. Thus, while in the last 30 years [p106] the total list of magistrates increased only by 10% and the number of officials of group 2 by 11%, the composition of the astynomes of group 3 grew by almost a quarter. Therefore, the existence of another 10--12 magistrates, as yet unknown to us, who carried out their duties in controlling the production of ceramic containers in Khersonesos at the very end 3c BC--the first decades of 2c BC, is not likely to cause any doubts. Taking this into account, there are full grounds to determine the length of the period embraced by chronological group 3 as 45--50 years.
The latter supposition is indirectly confirmed by the presence of 3 representatives of one family among the astynomes of this group: Hymnos son of Skythos, already known to us, Sopolis son of Hymnos, and Sopolis son of Sopolis (see Fig. 5). If Hymnos, as we saw, performed his duties somewhere within the limits of 215--210, then, taking into account a chronological break of 25 years, his son Sopolis should have become an astynome at the end of the 90s of the following century and the grandson, Sopolis son of Sopolis, in the mid 60s 2c BC. This very decade could be provisionally accepted as the end of ceramic stamping in Khersonesos.
The astynomes of chronological group 3 are preliminarily divided by us into 3 sequential sub-groups.
Sub-group A includes 11 astynomes, among whom 7 magistrates have, along with stamps [of type] III, impressions done according to late variations of type II. The remaining astynomes are represented by stamps only of the earlier variant 1 of type III. In the period embraced by sub-group A, the formation of the new type III of magistrate stamps takes place. In connection with this they are sufficiently varied typologically. The sub-group contains impressions of 8 variations. More than 80% of dies in the sub-group already produced impressions of flat shape. In the writing of the name of the magistracy all the stamps have the genitive of the participle. In the inscriptions of 80% of the dies one comes across lunate sigmas.
Sub-group B is exceptionally homogeneous. It includes 10 astynomes known almost exclusively by flat stamps of variation 2b of type III. In the readings of all the impressions the name of the magistracy is put in the genitive of the participle. A sharp reduction in the number of cursive forms of letters in the inscriptions is typical.
Sub-group B' is also typologically homogeneous. Out of its 12 astynomes 10 have stamps executed according to only 2 variations of variant 3 of type III. A die of the astynome Pythodotos son of Damokles made according to variation 3b' of type IV was a form of creativity on the part of one of the die-cutters.
Taking into account the fact that in each of the 3 sub-groups the number of astynomes known at present is practically equal, each of them can conjecturally be given a third of the period determined for chronological group 3. Sub-group A apparently fits fully into the 2 last decades of 3c BC; sub-group B embraces the first 15--20 years of the following century; sub-group B' end 80s--beg 60s 2c BC.
We realise that the proposed chronological classification of Khersonesan magistrate stamps cannot be considered final and needs further development and refinement. Undoubtedly, with the appearance of new impressions done by previously-unknown dies, a need to shift individual astynomes from earlier sub-groups and groups to later ones may arise. We should also take into consideration the fact that, due to the absence of a sufficient number of reliable archaeological complexes which contain stamps of chronological [p107]
groups 2 and 3, the absolute data of the borders of individual sub-groups as determined by us are of a preliminary nature (Fig. 7).
At the same time, as we see it, the proposed chronological system is more logical and more consistent typologically than the previous one. It makes it possible to provide a real base for clarifying such poorly-studied aspects of economic life in Hellenistic Khersonesos as the volume and the dynamics of production in pottery
The Chronological Groups of Khersonesan Magistrates Nr Name Type Dies Shape [table not reproduced]
Changes of Basic Features Characterizing Khersonesan Magistrate Stamps According to Chronological Groups. Group Dies(n) Shape Writing Monograms Cursive shapes of stamps(%) of Title abbreviations (%) of letters(%) [table not reproduced] List of abbreviations for Tables II and III: zh=grooved stamps, p=flat stamps, s/zh=weakly-grooved stamps, f=figured stamps, b/m=name of magistracy omitted, s=abbreviated from of magistracy, r.s.=magistracy in the form of genitive of noun, r.p.=magistracy in form of genitive of participle, i.p.=magistracy in form of nominative of participle.
workshops of tiles and ceramic containers in the city, the volume and dynamics of wine export in amphoras, as well as the place of the products of Khersonesan wine-makers in the markets of other centres of the Black Sea Area.
Moreover, the observed features of similarity in the evolution of astynome stamps in Khersonesos and Sinope make it possible to use the deductions received in studying Khersonesan impressions for making absolute dates of the chronological groups III--VI of Sinope more precise.