voyage, which varies with circumstances. They also agree that the land of the Seres with their metropolis lies to the north of the land of the Since, and that all that is further east than these is a Terra Incognita full of marshy lagoons in which great canes grow, and that so densely that people are able to cross the marshes by means of them. They tell
also that there is not only a road from those countries to Bactriana by the Stone Tower, but also a road to India which goes through Palimbothra. And the road from the metropolis of the Since to the port of Cattigara runs towards the south=west ; so the former city would appear not to fall on the meridian of Sera and Cattigaras, as Marinus will have it, but to lie further east."
" SERICE is bounded on the west by Scythia beyond Imaus, according to the line already defined (i.e., a line whose northern extremity is in long. 150°, N. lat. 63° and its southern extremity in long. 160°, H. lat. 35°); on the north, by the Terra Incognita, in the latitude of the Island of Thule ; on the east, by the Eastern Terra Incognita in the meridian of 180° from lat. 63° down to 3°0 ; on the south, by the remaining part of India beyond the Ganges along the parallel of 35° to the termination of that country in long. 173°, and then by the Sine along the same line till you reach the frontier of the Terra Incognita, as it has just been defined.'
" Serice is girdled round by the mountains named Anniba,2 by the eastermost part of the Auxacian Mountains, by the mountains called Asmireean, the easternmost part of the Kasian Mountains, by Mount Thagurus, by the most easterly part of the ranges called Hemodus and Serious, and by the chain of Ottorocorrhas. Two rivers of especial note flow through the greater part of Serice ; the river OEchordas is one of
these, one source of which is that set forth as flowing from the Auxacian range, and the other from the Asmirman range. . . . And the other is the river called Bautes, whieh has one source in the Kasian Mountains and another in the mountain of Ottorocorrhas.3
One might be reading the legislative definitions of the boundaries of an American state or an Australian colony. We see here how Ptolemy's Asiatic Geography was compiled. It is evident that he first drew his maps embodying all the information that he had procured, however vague and rough it might be. From these maps he then educed his tables of latitudes and longitudes and his systematic topography. The result is that everything assumes an appearance of exact definition ; and indications on the map which meant no more than " somewhere hereabouts is said to be such a country," become translated into a precision fit for an Act of Parliament.
2 I omit the latitudes and longitudes of the mountains, rivers, and cities named in this chapter.
3 There is, I suppose, no question that the Serice described here is mainly the basin of Chinese Turkestan, encompassed on three sides by lofty mountains. In Auxacia we probably trace the name of Aksu (De-guignes and D'Anville), in Kasia perhaps Kash gar (D'Anr.). Thy