Clay Seals on Letters  [see List Nos. 72-92]

The most typical example of documents that are sealed completely are letters. The letters were rolled up, tied up with a papyrus fibre and a seal was put in the middle of the roll, holding the ends of the papyrus fibre. The address was written to the right and/ or to the left of it. This was the usual practice for both Greek and demotic letters [see Plate].
 

Saltire Patterns on Letters  [see List Nos. 93-148]

In Roman times, a new practice emerges. On the spot where a seal is expected, between the address, a saltire pattern or cross (decussis  (44)) is found. It often escapes the attention of the editor that the saltire pattern is always interrupted in the middle. The interruption itself is very small. The only reasonable explanation for this is that a papyrus string was first tied around the letter, and then a saltire pattern drawn on the papyrus and over the string. So the parts of the cross that were not on the letter itself but on the string disappeared when the letter was opened (45).

To the right and/ or to the left of the pattern the address was written, e.g. List No. 99: Sarapivwni  X  tw/~ patriv ("To Sarapion, my father"). The pattern could not only split the address, but also words or names in the address, e.g. List No. 147: Plav  X  twni ("To Platon").

Few examples of this practice exist for the first century A.D. The saltire pattern is found mainly in the second and third century. There were two 'models' in use: a simple saltire pattern (type A1) and a saltire pattern with an extra line in the middle (type A2).
 
 

The purpose of this pattern is problematic. Some scholars believe that this group marks the place for the papyrus string. But as the pattern is drawn when the string is already tied around the letter, this explanation is clearly erroneous. It is also very unlikely that it marks the place for a seal, in other words that a seal was put over the inked cross (46).  That would imply that the sender of the letter first had to tie a papyrus string around the letter, draw the cross over the string and, finally, put a seal on the cross, which holds the ends of the tie. This is very improbable since seals and papyrus strings were attached to the papyrus in this way: first some clay was put on the papyrus. Then a papyrus string was tied around the papyrus and fastened in the clay. Finally some clay was put on top of it, in which the seal was impressed. Thus the papyrus string was held by the clay [see Plate]. With this system of sealing, it was not possible to ink a pattern over the string and on the papyrus.

Others assume, for different reasons, that the saltire pattern replaces the seal, e.g. Ziemann (47),  because it is found on the spot where one expects a seal; he believes the inked cross was used by the poor man instead of a seal. That the saltire pattern indeed replaced the seal, in other words that no clay seal was used, is confirmed by a few Roman letters recently found at Kellis in the Dakhla oasis (48).  They were still unopened and tied up with a papyrus string, over which a saltire pattern was inked. There were no traces of clay seals at all.

So the saltire pattern, which was drawn over the string, prevented tampering with the letter ? as did the seal ? since the letter could not be opened and closed without this being noticed (49).  A similar practice is attested with certainty for the Byzantine period (see below).
 

Elaborated Patterns on Letters  [see List Nos. 149-185]

A further development of the practice of the saltire pattern is found in Byzantine times, at least from the fourth century onwards. Amid the address is some kind of drawing, with parallel or crossed lines, but always interrupted by a relatively broad, rectangular space [see Plate]. Most editors call them ‘ornaments’. But they are not just embellishments. If one was to prolong those lines, it is clear that they would meet. So the writer closed his letter with a string or, more probably, a piece of linen, and drew a line-pattern on the papyrus and over the string. He usually did this on both visible sides of the rolled papyrus [see Figure 7], thus giving two patterns (50).  The letter writer normally uses the same pattern twice (51).  In most cases there is only one binding. Only No. 159 has been tied up in three places.

There are several types of patterns: parallel lines (Type B) and crossed lines (Type C). The lines can stand alone (Type B1 and C1) or they can fit in a rectangle or stand simply between two lines (Type B2 and C2). Finally, more original drawings, are found only once (Type D).
 
 

These drawings have the same function as seals and saltire patterns: to prevent the opening and re-closing of the letter without anyone noticing (52).  The same practice is found in Coptic papyri; the drawings are often wrongly called 'ornaments' by the editors (see List Nos. 182-185).