loading. Too heavy to be related to a ship ballast and not 
practical enough to be used for anchoring, they are furthermore 
located too far away from the edge to come from a coastal site. 
Regarding these points, there exists no historical nor 
archaeological source providing evidence of any occupation of 
the littoral in that area, in the Antique epoch, except a few 
sepultures (punt and round tiles) noticed in 1863 and 1895. 
The assumption that best accounts for these discoveries brings 
us back to the dredging of Marseilles’ harbour since the Roman 
epoch and until the modem ages. As it was the case with the 
horn of the antique harbour discovered in 1969 while the Stock 
Exchange Building was being built, the quays were probably 
made of materials stemming from Greek monuments that had 
been used as quarries during the whole Antiquity. 
The more recent searches of the Jules Verne and Ville-neuve- 
Bargemon squares enabled us to assess the great scope of 
dredgings during the Roman epoch in the Lacydon (old harbour 
of Marseilles) and at the same time, to find some vestiges of 
dredges that were abandoned in the 1 st and 2 nd century BC. 
First deposited in the “remblai hellénistique” of the Stock 
Exchange area, the sediments and debris resulting from the 
cleaning out of the harbour were subsequently removed further 
toward the open sea by the Romans, particularly in the Catalans 
area, located at the way out of the harbour. 
Well sheltered in windy weather, this area is well-known in the 
medieval and modem archives as the “quartier des infirmeries” 
or “quartier Saint Lambert”. It long constituted the privileged 
receptacle for deposits resulting from the dredgings that have 
been regularly repeated since the XVI th century. 
For this reason, the rich terracotta ware recorded around the 
blocks and fragments of statues belongs to three distinct periods 
: the Hellenic period, the late Antiquity and the modem epoch 
(XVI th - XVIII th centuries). 
During the search, a certain number of areas that were relatively 
rich in archaeological furniture, were spotted and then marked 
out with buoys by means of a flexible gridding used as a 
groundwork for the removal of the sludge performed with our 
suction pump. 
The items, meticulously excavated and in situ numbered, were 
then registered by photogrammetry. At the same time, a 3.4 
meter-deep sounding had been introduced down to the rocky 
substrate. It provides archaeologists with knowledge about the 
sedimentation and stratigraphy of the area. Besides, a series of 
aerial snapshots, triggered from an helium balloon, visually 
registers the position of relevant blocks indicated at the surfaced 
of the sea with numbered buoys. 
These elements were afterwards all more precisely registered by 
means of a GPS receiver. 
Finally, the whole site, which altogether covers an area of 
several hectares, was prospected with an echosounder for 
sediment layers, which is capable of detecting buried blocks 
under the upper sediment layer. 
The significant number of anomalies that has been recorded up 
to now have not been checked on the field yet, which allows us 
to imagine the richness of the area. 
Figure 3. Interpoled grid based on sonar measurement. 
< J ^ i&? vi 
4. THE SURVEY METHOD 
4.1 Sonar 
We used a multibeam sonar, GeoSwath with a 250 Khz 
frequency, set up by Maritech company 1 . This sonar is designed 
for swallow water (up to a depth of 200 meters) and was