84 ‘Mr. Bell’s Remarks on the ~ 
other method of taking longitudes was adopted by them than was 
used by the geographer of Alexandria. The fact is, the Oriental: 
‘Tables are almost mere transcripts. of each other, and those of: 
Ulugh Beigh are evidently transcribed from those of Nasroddin. 
For instance, the cities in Mawalnahar, laid down in both tables,. 
agree punctually, except in the case of Samarcand, wherein they. 
differ 23 minutes, Nasroddin placing it in 40° N. lat. instead of- 
39° 17’, as given in Ulugh Beigh. Now as we know that the lat-' 
ter determined the latitude of his own capital, we may be pretty- 
confident that its latitude is correct, but we cannot depend on any 
of the rest. Can we conceive that Samarcand was the only faulty 
latitude in that part of Nasroddin’s tables? The bare fact of the: 
disagreement in this solitary instance of Samarcand, and their punc- 
tual agreement in all the rest, is a plain proof that the latitude of 
Samarcand was the only one observed by Ulugh Beigh in all his 
dominions, and that if he had taken the latitudes of the other cities, 
as Khojund, Otrar, Tashkunt, Tonkat, Esfijab, Taraz, Aksikhat, 
Andecand, Khowakund or Kokaun, Balk, Bokhara, Zarnuck, Soga- 
nian, Termid, and others, his table could not possibly have agreed 
so well with that of Nasroddin. Tashkunt we do know to be 33 
minutes too far S. in the tables of Nasroddin, Abulfeda, and Ulugh 
Beigh, which make it 42° 30’ N. lat. whilst in the great map of 
China, made from observation, its true latitude is 43° 3’ N. By 
the same tables, Cashmere is placed 3° W. of its true position. __ 
Rennel admits that the positions of Candshar, Caubul, and Mul- 
taun, are erroneously set down in these tables, and that in the 
whole southern line, between the Caspian Sea and Lahore, through 
Heraut, the numbers are either falsely written, or erroneously cal- 
culated ; and how therefore can we depend upon them in any line? 
He admits that the longitudes of these tables were calculated to 
the computed distance,—a very uncertain mode of fixing either 
longitudes or latitudes ; for unless we know the nature of the in- 
tervening space, the degree of winding of the roads, and what pro- 
portion of sinuosity should be allowed, how can we tell what is the _ 
longitude or latitude of any place? For instance, we have a strik- 
ing proof in the case of Cashgar, how little dependence can be 
placed on itineraries to fix the latitude of a place. Rennel having 
placed Cashgar in 42° 45’ N. lat. by itineraries from Bokhara, 
Tashkunt, and Samarcand, and fixed its longitude at 71° 44'S. of 
Greenwich, endeavours to confirm the same by the route from Cash- 
mere through Little Tibet, given in the travels of Bornier. This 
gentleman states from information, that 44 day’s journey of a cara- 
van are allowed for the intervening space. Rennel allows 12 geo- 
graphical miles of direct daily travelling to accomplish this, and 
accordingly finds the latitude of Cashgar to be 42° 45’, or 528 geo- 
graphical miles of direct distance from Cashmere. Now, Cashgar 
is found by celestial observations to be only 5°, or 300 geographical 
miles N. of Cashmere, instead of 8° 25’ by the method he employ- 
ed. This being the case, the daily rate of caravan travelling across 
