Ceramic stamps containing one or another representation alongside a reading are, in the same way as coins, seals, and other analogous archaeological monuments, a valuable historical source. The comparative method used as one of the main methods of studying coins is fully applicable to the study of ceramic stamps as well. The diversity of ceramic dies is one of the characteristic features of the stamps from Sinope. Even stamps with the name of the same astynome often differ among themselves. These differences have been frequently noted by all scholars of Sinopean stamps. Already P.V. Bekker noticed that stamps with the name of the astynome Histieios and an eagle on a dolphin differ in the position of the device and in the supplementary monograms on some of them.  V.V. Shkorpil frequently observed a difference in the arrangement of the lines of the inscription,  a shift of device on stamps with the same readings  and even the substitution of one device for another.  Giving a commentary on one of the stamps, V.V. Shkorpil noted that quite different devices are impressed beside the same inscriptions.  Iu.N. Iurgevich pointed out a difference in the dies connected with some astynomes.  E.M. Pridik noticed more than once different devices for the same astynome,  differing treatment of the astynomic device, and so forth.
B.N. Grakov also points out how many there are of such cases and adduces some further characteristic examples of the differing treatment of the device of the astynome.  He considers that changes in the details of the devices lie in the hands of the potters, the authors of the stamps, who varied the astynomic device in their own way, retaining only the basic motif  and even sometimes exchanging one device for another. [p57]
V.V. Borisova notes that, in the study of the stamps of the same Khersonesan astynome, dies have come to light which have certain differences. However, she explains this fact by saying that one astynome did not manage to stamp the whole ceramic production due to a large demand for containers and so other astynomes also engaged in the stamping, members of a board, who stamped containers and tiles with the name of the first astynome. 
Fig. 1 Fig. 2
Thus, in the first case the differences in the dies depends on the potters, in the second, on the astynomes.
The application of the comparative method, as it seems to us, permits us to make the most correct decision in the question of the difference in the types of the dies, to explain the reasons on which the difference depends, and to demonstrate the results of that dependency. Let us adduce some examples.
Examining the stamps with the name of the astynome Mikrias,
we notice that one group of stamps has the reading
The second group of stamps has the reading
Fig. 4 Fig. 5knows a stamp with [the name] of the potter Histieios.  Thus there are known six stamps, constructed according to the above-mentioned formula (Figs 4, 5).
On comparison of both variants a difference is noted in the arrangement of the component parts of the reading.
[Fig. 6 in margin]In the first case at the beginning stands the name of the astynome, then the title of the magistracy; in the second on the other hand, there is first the magistracy, then the name of the astynome. There is a difference also in the treatment of the device.
[Fig. 7 in margin]In 20 cases the figure of a horse is very strongly represented to the left, rearing, with a beautifully curving tail. In three cases the figure of the horse is small, he walks quietly (or stands) to the right. The two variants differ also in the letter forms and in the forms of the genitive case of the name of the astynome.
The 20 stamps mentioned above of the astynome Mikrias with different names of potters have a wholly similar type of die; they are engraved by the same hand, on all of them there is a single treatment of the device.
[Fig. 8 in margin]Three other stamps of the same astynome have another type of die, common for all three stamps with different names of potters. The handwriting and the similar treatment of the device also unite them.
The difference in the treatment of the device, then, cannot be made to depend on the potters:
[Fig. 9 in margin]20 different potters ought to have 20 different treatments of the same device, and we have one treatment common to 20 potters, and another treatment common to three potters.
Let us look at one more example. In the group of stamps with the name of the astynome Aischinos  three basic variants stand out (Figs 6, 7, 8).
The attribution of all the stamps to one magistrate does not admit of doubt: a single device unites them, a club. However, in comparing the stamps among themselves a difference is noted in the dies.
[Fig. 10 in margin]In the first case the structure of the reading recalls the first variant of the above-mentioned example of stamps with the name of the astynome Mikrias.
[Fig. 11 in margin]The magistracy
[Fig. 12 in margin]The second variant differs sharply from the preceding and recalls the second variant of stamps with the name of the astynome Mikrias. The magistracy
[Fig. 13 in margin]A third variant does not differ from the second in the formula of construction, however the club is here placed with its thickening end upward. In addition, the two last variants have different letter forms. There are eight stamps of this variant know, with different names of potters. Thus, in the three variants of stamps with the name of the astynome Aischinos which differ basically in the different position of the astynome device, we come across the names of 14 different potters.
Let us adduce an example of the variation of dies with the name of yet another astynome --- Kallisthenes (Figs 9, 10, 11).
[Fig. 14 in margin]
The attribution of all three variants to one astynome does not allow objection. The second and third differ from the first in their devices. In the first variant there is a representation of Nike, galloping in a chariot. There are known four such stamps with the names of four different potters. In the second variant the device is the prow of a ship (the names of two potters are known); in the third --- a bunch of grapes (one potter). Here, as also in the preceding examples, the potters are distributed according to several variants of stamps, however the basic features of these variants are not the different treatment of the same device, and not its position on the stamp, but different devices for the same reading.
The grouping of potters around the different treatment of the device or around different devices for a single astynome may, as it seems to us, be made to depend only on the die-cutters. At the time of the astynome Kallisthenes there were working, probably, not fewer than three die-cutters of ceramic dies; in filling the orders of certain potters, they accompanied the stamps with different devices, sometimes ignoring the official device of the astynome.
We have adduced only three examples from the groups, of various times, of ceramic dies. It is possible to adduce a large quantity of such examples from all the chronological groups (B.N. Grakov numbers six groups of astynomic stamps). Variants of types are especially noticeable in those cases where astynome groups are represented most fully (up to ten or more potters).
It is interesting to remark that, in the limits of one chronological group and in the adjacent chronological groups it is possible to notice, on comparing the dies, the work of one die-cutter for various astynomes (Figs 12, 13, 14).
The first two stamps may be assigned to the third chronological group of B.N. Grakov,  the last to the fourth group.  Thus, the difference in the variants of the types of stamps depends not on the potters, as was proved above, and not on the astynomes. The last example with the names of three different astynomes, but with a similar formula of construction of the reading, confirms this.
Difference in the types of dies depends on the die-cutters, who filling the orders of different groups of potters and striving to distinguish their own production from the production of other cutters, each in his own way, treated the device that accompanied the reading and positioned it on the stamp, or changed the formula [p60]
[Fig. 15 in margin]of the construction of the reading or used different devices for the same astynome. We are no longer speaking of the different handwritings and different types of letter forms used by te die-cutters.
[Fig. 16 in margin]Comparison of the dies leads to the conclusion that, for any astynome, it is possible to distinguish several variants of types of stamps, which will have among them certain differences. These variants always unite whole groups of potters. It is possible to trace them in all the chronological groups of B.N. Grakov.
The Kerch museum has a stamp
[Fig. 17 in margin]At first the cutter intended to put the name Philokrates, but, before he had finished it, he stopped, leaving it at
[Fig. 18 in margin]Then, above the inscription, he engraved
Observations accumulated in the study of the stamps within astynomic groups permit us to pass on to the comparison of the actual astynome groups themselves.
[Fig. 19 in margin]These comparisons lead to the establishment of a sequence in the changes which occur on Sinopean stamps. P.V. Becker was one of the first to point out this sequence.  After him followed B.N. Grakov,  who made the palaeographic method and the method of establishing the synchronicity of the names of the magistrates and potters the basis of his investigation.
In its development from the simplest form in the shape of a single name of a potter in the genitive case and a device of the town's symbol (Fig. 15) to the complex form, which consisted of the title of the magistracy, the name of the astynome with his patronymic, a device for the astynome and the name of the potter in the nominative case (Fig. 16), Sinopean stamps underwent a series of changes, these changes affecting both parts of the stamp --- readings as well as devices.
Comparing the stamps of different astynomes, close in time, one with
another, we may notice how gradually the readings changed. Let us
compare, for example, stamps with the names of astynomes of the first
chronological group of B.N. Grakov (Figs 17--19). For the same content
of device (city symbol) the reading changes. In the first case there
is as yet no title of the magistracy, then it appears in a most
abbreviated form and later it acquires the form
Changes in the reading show themselves distinctly in the comparison of some astynome groups within the limits of a definite segment of time. These changes [p61]
Figs 20--28consist of the following: a) the composition of the reading, b) the form of the title of the magistracy, b') the formula of construction, g) the use of the cases of the names of the potters. The changes in the reading takes place so consecutively and in such a regular order that they can be broken down into stages, each of which constitutes a definite step in the development of the reading. Let us examine each of the changes mentioned above in general outline.
a) The change in the composition of the reading consists in the fact that at first on Sinopean stamps we see only a single name of a potter (Fig. 20), then the name of the astynome appears (Fig. 21), which soon receives the title of the magistracy accompanying it (Fig. 22).
At a certain period of time there appears and disappears the patronymic of the potter (Fig. 23).
The last change consisted of the obligatory use of the patronymic of the astynome  (Fig. 24). [p62]
b) The change of the form of the title of the magistracy.
The title of the magistracy accompanying the name of the astynome at
first was expressed very briefly ---
[Fig. 29 in margin]
Afterward it was changed and for a long time
At a certain stage
[Fig. 30 in margin]
It should be noted that at the end of the early group the title of
the magistracy is expressed sometimes by the full genitive case, but it
was not decisive for this time, just like the use of the participle in
some stamps with the inscription on the sides of the rectangle.
Following on the early group is a quite significant group of stamps with
potters' symbols which knows neither the full genitive case of the title
of the magistracy nor the participle with the name of the astynome. Only
b') Changes in the formula of construction (of the place of the title of the magistracy on the stamp). Application of the comparative method permits us to make one more observation about the reading on Sinopean stamps. P.V. Becker was one of the first to classify three basic variants of the schemes of construction of the inscription. 
B.N. Grakov, basically accepting the classification of P.V. Becker,
introduced a more exact distribution of the stamps according to the
variants of construction of the reading. To the first variant he
assigned those stamps in which the reading begins with the title of
the magistrate, to the second, those in which the title of the
magistracy stands between the proper names, to the third, those in
which the name of the potter precedes the name of the astynome. The
group of stamps with
Poseidei/ouof tou= `Hfaistodw/rouclub up two a'stuno/(mou)readings] `Ikesi/ou tou=cluster Dionusi/ou kerame/ws
However the variants suggested by B.N. Grakov of construction of the
reading do not exhaust all the variations of the formulas according to
which Sinopean stamps are built, and, on the contrary, some of them
embrace an insignificant portion of the stamps and can be subsumed
under basic formulas. Thus, the second variant of B.N. Grakov --- with
the magistracy in the middle of the inscription --- does not make the
placing of the names of the astynome and potter on the stamps precise,
while the fourth variant, where stamps with
Figs 31--41of construction of the reading, in which is taken into account both the position of the title of the magistracy and the place of the names of the astynomes and potters.
Scheme 1. Stamps without the magistracy (Figs 29 and 30).
Scheme 2. Magistracy placed between the proper names with the name of the astynome preceding (Figs 31, 32, 33, 34) --- second variant of B.N. Grakov.
Scheme 3. Magistracy placed between the proper names with the name of the potter preceding (second variant of B.N. Grakov; Figs 35, 36).
Scheme 4. The magistracy stands at the beginning of the reading (first variant of B.N. Grakov; Figs 37, 38).[p64]
Scheme 5. Magistracy at the end of the reading (third variant of B.N. Grakov; Figs 39, 40, 41).
[Fig. 42 in margin]
Scheme 6. Distribution of the basic parts of the reading onto two dies (fifth variant of B.N. Grakov).
Another variant of this scheme is possible, when in the
first place there stands the magistracy
and the word
These schemes include all the possible variants of the formulas of construction, however, they are built without regard to the chronological succession of the change of reading and device.
[Fig. 43 in margin]
g) Changes in the use of the cases of the names of the potters. About half of all the astynome stamps have the genitive of the name of the potter (Fig. 42), which is replaced by the nominative case (fig. 43).
[Fig. 44 in margin]
In early groups, for which the use of the name of the potter in the genitive case is characteristic, the nominative case is sometimes to be encountered. In this connection stamp nr 5179 of the Kerch museum is interesting (Fig. 44).
[Fig. 45 in margin]
[Fig. 46 in margin]In the meantime in a list of astynomes of the first B group of B.N. Grakov there is the name of the astynome Khabrias,  to whom the above-mentioned stamp should also be assigned.
Sometimes the use of the name of a potter in the genitive case is to be met with also in much later groups, in which the nominative case predominates, but these appearances are isolated. By and large the changes of the reading present a rather ordered picture and the place of putting of the title of the magistracy in front of the name of the astynome or after it, at the end or at the beginning of the reading is not accidental.
[Fig. 47 in margin]The devices also change. However if we are to follow up the changes of the reading, comparing groups of astynomes among themselves, then the changes in the devices are not detected. For that one must compare chronological groups.
Stamps with early astynomes, which B.N. Grakov introduced into the first chronological group,  are accompanied by an eagle on a dolphin. [p65]
B.N. Grakov proposed that we consider this device astynomic, since with one astynome, for example Histieios,  independently from the names of the potters, there is invariably an eagle7 on a dolphin.
[Fig. 48 in margin]However this device is peculiar not to one, but to several astynomes, close in time, and therefore there is no basis for considering it to be the symbol of a particular astynome. In addition, the eagle on a dolphin is represented also on stamps which do not have the name of an astynome, and which contain only the single name of the potter.
[Fig. 49 in margin]At the present time there are known 12 stamps of this type. Therefore the device of an eagle on a dolphin cannot belong either to one astynome or to a given potter. (Figs 45, 46, 47).
This device is considered to be the city symbol of Sinope, her emblem,  and obviously must, for contemporaries, have pointed to the attribution of the amphoras with this stamp to this particular centre.
[Fig. 50 in margin]The stamps of Rhodes provide an analogous picture, where there occur representations of a lotus flower and a head of Helios.
[Fig. 51 in margin]M. Nilsson considered that these two symbols indicated the Rhodian origin of the goods, just as did the inscriptions
The use of the city symbol constitutes a definite period in the development of the Sinopean stamps. It embraces all the astynomes of the first chronological group of B.N. Grakov who use this symbol.
[Fig. 52 in margin]
The city symbol is replaced by an unusual assortment of devices. To whom do these devices belong, what is their content? Let us look at several stamps of the astynome Aischinos  (Figs 48, 49, 50).
All the stamps belong to one astynome; however, on all three stamps the devices are different. Since on the stamps are placed the names of different potters, the devices may be considered as symbols of these potters.
[Fig. 53 in margin]
The use of potters' symbols is an indication of the second period in the development of Sinopean stamps. Here should be assigned stamps of the first B, and partially of the second, chronological group of B.N. Grakov, particularly those about which the latter writes that their alphabet is close to the first group. 
Comparison of the dies shows that the selection of the symbol in the second period did not depend on the potter; the same potter with different astynomes [p66] had different devices. For example, the potter Nikomachos with the astynome Aischinos had as his device the upside-down head of a gryphon (Fig. 51), while with the astynome Mnesios he had an owl (Fig. 52).
At the same time individual devices are to be found on stamps with different names of potters (Figs 52 and 53).
This phenomenon may depend on the die-cutters, inasmuch as the treatment of the devices, for example the lotus flower, is so similar for different potters that it produces the impression of having been imprinted by one die.
For a certain period of time Sinopean stamps are expanded with yet another device. Thus, comparison of the dies of the astynome Epielpos shows that stamps for this official have a varying number of devices (Figs 54, 55).
Fig. 54 Fig. 55
Once they have appeared, the second devices quickly become permanent for a given astynome. There begins the period of the existence of the astynomic device. However, the appearance of this device does not at first entail the disappearance of the potters' symbols. On the contrary, potters' symbols, just like the astynomic devices, become, in contrast to the preceding period, to some degree permanent.
The existence of several devices on a stamp has been attributed by scholars to various causes. P.V. Becker wrote in this connection that most probably one of the two devices referred to the astynome, the other to the fabricant.  B.N. Grakov considered this phenomenon to be fortuitous. "Sometimes," he wrote, "if the positioning of the inscription left large gaps beyond the place for the basic device, the die-cutter filled them with small supplementary devices." 
Let us see whether these supplementary devices appear to be fortuitous. Let us take the potter Poseidonios from the second chronological group of B.N. Grakov and let us consider a table of the coincidences of the name of this potter with different astynomes on the ceramic stamps.
Astynome name Potter name Devices Heronymos Poseidonios flower Kallistratos " " Protagoras " " , hand Isokrates " " , amphora Theogetes " " , bird Aristoboulos " " , Hermes Herakleides " " , lion Aristokles " " , bull Polycharmos " " , kantharos Eucharistos " " , lion, man's head Hikesios " " , elephant Artemidoros " " , cluster
From the table it is clear that for different astynomes there is
repeated the same device, the lotus flower, which is undoubtedly connected
with the potter Poseidonios.
The second devices belong to the astynome, with whose names [p67] the
name of Poseidonios is found. For example, on a stamp with the names of
the astynome Aristoboulos and the potter Poseidonios there occur two
devices. One of them, the head of Hermes in a petas
But in the same group there are stamps with the name of the potter Poseidonios with whom other devices are repeated.
Astynome's name Potter's name Devices Kallistratos Poseidonios cluster and ear of grain Epielpos " " Theogetes " " Apollodoros " " Artemidoros " " Aristoboulos " " , Hermes Aristophanes " " , lion's head Gyrittos " " , bull's head
Here with different astynomes there are repeated not one but two devices, cluster and ear of grain. In these two devices we must see the symbol (symbols) of the potter Poseidonios 2. The third devices, like, for instance, the head of Hermes with Aristoboulos, belong probably to astynomes. In the Kerch museum there is a stamp with only one name and
Fig. 56 Fig. 57 Fig. 58the patronymic of the potter and the devices cluster and ear of grain, which undoubtedly belongs to Poseidonios 2 (fig. 56).
Let us take yet one more example with the name of the potter Hephaistios.
Astynome's name Potter's name Devices Heronymos Hephaistios cluster and kantharos Epielpos " " Apollodoros " " , broken off Herakleitos " " , lion Aristokles " " , bull Gyrittos " " , bull's head Polycharmos " " , broken off Iphis " " , cluster and ear of grain
Here in connection with different astynomes there are repeated the devices cluster and kantharos, while, beginning with Apollodoros, these two devices always occur to the left of the inscription (Fig. 57). They are the symbols of Hephaistios. Thus the plurality of devices is not an accidental phenomenon and not the filling of empty spaces on the stamp, but the regular normal distribution of the astynomic and potter symbols. The potters Poseidonios 2 and Hephaistios have two devices each, and the quantity of devices on the stamps is therefore increased to [p68] three, and in one case even to four, since the astynome Iphis, assigned by B.N. Grakov to the third chronological group,  probably also has two devices (Fig. 58).
The discovery of an amphora with both handles preserved in the Zelenski tumulus also attests the simultaneous existence of devices of astynome and potter.  On one handle is impressed the name and patronymic of the astynome Poseideios, accompanied by a device, a club. On the other handle appears the name of the potter Hikesios with patronymic and device, a cluster of grapes. Here, as also in the preceding two examples, both astynome and potter have their personal devices; this circumstance is underlined by the presence of independent dies for both people.
The time when two or more devices, some of which belong to the astynome and others to the potter, occur on the stamps is the third period of the development of Sinopean stamps. To it should be assigned the greater part of the second chronological group of B.N. Grakov, and a few names of astynomes of his third chronological group. One must note that precisely at this period there occurs a transition from the abbreviated form of the title of the magistracy to the full form, expressed in the genitive case. The complex from the Zelenskii tumulus lies wholly and fully within the limits of the third period of Sinopean stamps.
After that the potters' symbols on the stamps disappear in the same way as in its own time the city symbol---an eagle on a dolphin--- disappeared. As a result of all the changes only one astynomic device, which is the distinctive feature of the final, fourth period in the development of Sinopean stamps, remains on the stamp. It embraces the third, fourth, fifth and sixth chronological groups of B.N. Grakov (Figs 1--16, 23, 24, 28, 33--38, 41--43). The periods and stages of development may be placed into a definite chronological framework which will allow us to define more precisely not only the chronological groups themselves but also the distribution of the astynomes within the chronological groups.
The general scheme of the development of Sinopean stamps, taking into consideration the changes occurring in the readings and devices, is visualized as follows:
This gradual evolution of the Sinopean stamp from a simple type to a complex one, which was taking place sequentially within the limits of certain periods, allows us to outline (at least approximately) their relative chronology. It is true that in the process of development one sometimes observes a return to old forms, but not for long, and moreover this recurrence of old schemes could not shake the general picture of the gradual development of the types of Sinopean stamps which was outlined above.
Fig Reading Devices Inv nr Where published* [table omitted]