J. H. Young, Apollo, Altar II Apotheke Record (Ayia Anna) 1938, 1947John Howard Young

10. Nov6-7, 1938 I spent the morning in assorting out the heads more carefully, following the same general plan but grouping some more carefully according to beards and color and separating one whole class which has a headdress falling down over the shoulders and thus shows both on head and body. In the afternoon I made an experimental mold of a face from wax and cast it in plaster. I shall wait until the best process of moldin casting is discovered and then record the formula in the mold section. The purpose of a plaster cast of each mold is to facilitate recognition for the catalogue and have a head which may be numbered and carried about continually, and which shows the face as in the original mold, without any trimming or head-dress.

Nov. 7Today continued the separation of the various types of fragments, with the same system as was used above. With the use of a large table it was possible to put all the heads out in the same room. At two o-clock the separating was finished, and the two boys were set to finding joins. By five we had definitely proved that the system worked after all, and that, as suspected, the eight sections

Right page

Nov. 7-8, 1938 11. were entirely mixed together and for purposes of joining had no significance. Of the 19 joins found, not one join was made within a section and eight within one layer were overshadowed by eleven between layers. The significance of this must be later discussed. Meanwhile sectional provenience of each piece member of the join is recorded on a ticket and the pieces go to Christo to be glued in place and then to my desk. The baskets of unwashed terracottas were given to Nikolaos, one by one, for washing. Christos has begun an examination of the stone pieces which have been turned over to his section from the sorting process. Number of joins - 19

Nov. 8Work as usual today, Christos working on the stone pieces and gluing the heads, the two boys searching for joins, and Nikos washing. I am forced to consider the problem of what is to happen to the joined pieces. I see no point in numbering pieces which will later be joined together, causing a great deal of complicated re-cataloguing. Therefore I devise the following method: a horseman group (all that we have had so far) is composed of three elements which might conceivably be considered the basis for a group and so numbered: the horseman's head; the horseman's body-trunk;