By the Rev. E. H. Goddard. 409 
Another semicircular bath occurs in the Roman villa at North 
Leigh, Oxon. See Morgan’s Romano-British Mosaic Pavements, 
p- 118; and in 188], at Wingham, in Kent, a bath with walls and 
floor of mosaic was discovered, Jdid, p. 151. 
More than one fountain of semicircular shape, with coved head, ” 
and lined with mosaic, have been found in a perfect state at Pompeii. 
The tesseree at Box are of three colours—the outside groundwork 
cream-coloured, the borders of the central panels and the Laby-’ 
rinthine fret which forms the pattern, a dark slaty blue, and the 
groundwork of the pattern itself an ivory-white, a good deal lighter 
than the cream of the outside margins. 
All three of these seem to be formed of very hard close-grained 
limestones, capable of taking a high polish. Some of the cream- 
coloured tesserze, for which I am indebted to Mr. Stier’s kindness, 
have been submitted to Mr. F. W. Rudler, of the Jermyn Street 
Geological Museum, who writes as follows :— 
“The two tesserse which you have sent are similar to others which I have seen 
from Roman pavements at Cirencester and at Silchester.—They are a fine-grained 
limestone such as might well be derived from some of our secondary strata. I 
have submitted your specimens to my colleague, Mr. H. B. Woodward, who has 
made a special study of all the Jurassic and Triassic rocks of Great Britain, and 
he is satisfied that the rock of your tessere might be obtained either from the 
White Lias (Rheztic) or more probably from the White Limestone of the Great 
Oolite, which is well developed near Cirencester and was used according to the 
late Professor Buckman for Roman mosaic pavements in that neighbourhood.” 
{In Buckman and Newmarch’s Remains of Roman Art in Cirencester 
it is stated that the hard white or cream-coloured stone used for 
tesserze at Cirencester, which was formerly supposed to be “ Palom- 
bino ” marble from Italy, is really from a compact bed of fine-grained 
stone in the freestone quarries round Cirencester itself from the 
middle of the Great Oolite.] 
This Box pavement is not mentioned amongst the Wilts examples 
in Morgan’s Romano- British Mosaic Pavements, nor has it, so far as 
I can discover, been noted elsewhere. Its date, from coins said to 
have been found with it, is probably the fourth century. It has, as 
will be seen from the facts mentioned above, evidently formed part 
of a considerable villa. 
