fragrant cups of mocha to those parched by a well-nigh Arabian sun! It does not seem that any Roman remains had been discovered in this immediate neighbourhood before,[233] though the appearance of some of the other blocks of masonry made it probable that this was not the only classic monument concealed in the foundations of the Mahometan Han and its out-buildings. One would, indeed, expect to find Roman remains in this vicinity, not holy from the fact that in the neighbouring mountains about Foinica, and again to the east about Vareš, are traces of Roman gold-mines, and that a Roman road has been traced tending from Salona towards these centres of mineral wealth; but for another and still more cogent reason. This is the presence of hot-springs, which we passed to the right of us only a little way further on the road, at a place called Illidzje. The Romans, with their usual instincts, tracked out these natural baths among the Illyrian wilds; and Roman remains in Bosnia, when not connected with mining enterprise, seem especially to centre round such spas. At Novipazar,[234] in the province of Rascia, the sulphur springs still bubble up into an octagonal marble basin near eight yards in diameter, described as of Roman workmanship, and are still sheltered from the elements by an octagonal chamber, supporting a cupola, which also dates from the days of the Cæsars. At Banjaluka[235] on the Verbas the hot-springs are still housed in a similar edifice; and the curious may survey modern Mahometans taking their enjoyment in one of these very baths which have supplied the prototype of the early Christian baptistery. The name of Banjaluka[236] itself preserves a Roman element, _Banja_ being no other than the Latin _balnea_. The Sclavonic settlers borrowed the word from the earlier Roman population of these lands—thus witnessing to the purely Roman associations of such spas; and so completely has the word passed into their language that _Banja_ at the present day is applied by the Southern Sclaves to all hot-springs and baths. But though not necessarily, therefore, proof positive of a Roman connection, this derivative, when applied to local names, may perhaps afford presumptive evidence that the virtues of such spots were not unknown to the Romans. It is at least worth noticing that the thermal springs at Illidzje, not far from the place where we found the bas-relief of Cupid, are still known to the Bosniacs as ‘Banja.’