186 ON THE ANCIENT GEOGRAPHY 



his view of the situation and geographic relations of Serica it- 

 self. 



The first proof to which we shall attend is that derived from 

 his longitudes. These are universally allowed to carry Serica 

 very far beyond Thibet or Eastern Tartary. M. Gosselin, 

 however, insists, that as they carry it beyond China also, and 

 into the very heart of the Pacific, no regard whatever can be 

 paid to them. That eminent writer, however, has here re- 

 markably overlooked his own important illustrations of the 

 mathematical geography of Ptolemy. He has shewn, that the 

 errors of that ancient writer, whether they reside in the value 

 of the itinerary stadium, or, as I rather suspect, in the measure- 

 ment of the great circle of the earth, are not random errors ; 

 that they proceed in a regular train, and that there exist data 

 by which they may be calculated. It appears, that by redu- 

 cing his longitudes in the relation of seven to five, they will 

 all approximate pretty nearly to the truth. Now the remotest 

 longitude given in Serica is 180 degrees east from the suppo- 

 sed meridian of the Fortunate Islands, which was fixed 2\ de- 

 grees to the west of Cape St Vincent 180°, reduced in the 

 above ratio, will give 128° 30', or 117° from the meridian of 

 London, which is three or four within the eastern frontier of 

 China. It thus falls short of the ocean, which will account for 

 the only circumstance ascribed to Serica which does not agree 

 with China, that of its being bounded on the east by unknown 

 lands. The truth is, a bounding terra incognita was quite a 

 theory of the school of Marinus, which they everywhere made 

 to succeed to the original theory of a circumambient ocean. 

 If the narrow peninsula of Malacca was considered as stretch- 

 in & into an unknown extent of continent, much more might 

 this be supposed of an empire so vast as China, and of which 

 the interior was so little explored. But Mela, Pliny, and 



in 



