Hottest day of the season so far at Aragatsiberd--we broke more frequently for water and rest, but still managed to get a lot done in AB3, AB4, and preparing to map the Aragatsiberd surface architecture.In AB4 we changed to locus 2 as the locus 1 sod was completely removed early in the morning. Locus 2 consists only of the lower, more easterly 5 meters of the trench. Its western edge is very soft and rock-free, but as one moves east the matrix become increasingly rocky.In AB3 we continued to descend by trowel in the locus 9 quadrant. We also worked with trowels and shovels in a new locus 11 northwest quadrant, direcly south of wall locus 4. By the end of day we had produced additional wall courses in locus 11 and exposed a good deal of ceramics and a possible groundstone wedged below the locus 4 wall in locus 9. We took C14 samples from both loci today. At present all three major wall sections appear to intersect, making it difficult (without a floor or other orienting surface) to determine which is earlier and later. This is only complicated by the fact that only Lchasehn-Metsamor materials appears to be present in the wall-abutting matrixes. No EB...We also spent some time today searching the hill for surface architecture. We plotted all of the visible stone architecture by hand on the new topo printout. We will soon shoot in the architecture and plot it digitally.