THE EXCAVATIONS 55 



passing througli tliat entrance. On opening up the 

 southern gateway tlie road leading from that entrance to 

 the central building was also found to be in excellent 

 preservation. The present surface of this road is practic- 

 ally level, and the clay subsoil on which the foundations 

 rest seems also to have been worked level, both being 

 devoid of the usual camber or curvature. The road is 

 about 1 ft. 3 in. thick, and is composed of large rounded 

 stones, smaller cobbles, pebbles, and coarse gravel. The 

 whole of these have been well rammed together and 

 thoroughly consolidated. As neither camber nor wheel 

 ruts can be detected, it is possible that the present surface 

 does not represent the upper surface of the original road. 



DKAINS. 



The investigation of the Roman drains is rendered more 

 difficult by the fact that the site was drained in the last 

 century at the time of the cotton famine. Before 1905 

 one Eoman drain had been uncovered, which is shown in 

 the plan as pursuing a somewhat irregular course north- 

 wards towards the N.W. corner of the area. This was 

 traced back last summer to the southern side of the Via 

 Princiimlis, where it was lost. Two other drains have 

 since been discovered. The first was found to terminate 

 in the rampart wall near the north-east comer, and is so 

 marked on the plan. It has not yet been opened up. The 

 other runs parallel to the Via Principalis about half-way 

 between that road and the south wall, and has been fol- 

 lowed practically as far as the central building. It is 

 formed of large flags, but has apparently been narrowed 

 by lateral earth-pressure. The clayey subsoil of the site 

 causes it to hold much water, and even in the summer 

 excavation is somewhat impeded for this reason. 



