Ancient Linear Measures. 221 



In Petrie's " Inductive Metrology," p. 129, there are given no 

 less than fifteen varieties of the so-called gaz, gutze, or cubit, 

 varying in length from 14*9 to 38*3 inches. In more modern 

 times the \ gaz is also common =s 6-93 inches. The hasta Mr, 

 Petrie considers to be a very ancient Aryan measure, sometimes 

 found in Greece and Asia Minor as = 17*9 inches ; but in India 

 it may be reckoned at about 18*4 inches. Another unit is 11-63 

 inches, which may, perhaps, be referred to the Eoman foot in 

 Mohammedan buildings in Turkey and Persia, where the Greek 

 or Poman Empire extended. 



Mr. James Fergusson informs me that no ancient buildings of 

 India are set out with sufficient exactness to recover a measure 

 from them, which may even apply to buildings of the time of the 

 Mogul Empire, when Europeans were employed. It is nearly 

 impossible to ascertain the length even of the Illahi gaz, and it 

 might almost seem that the Hindoos never employed any other 

 " rule " than the cubit or forearm of the reigning king. Mr. 

 Petrie, however, gives the Illahi gaz = 34*1 inches English == 

 41 digits ; and it has probably nothing to do with the digit of 

 the early Hindus, which is connected with the hasta, though it 

 would come to about the double of it. 



It is difficult to suppose that the old Assyrian cubits of 19-04 

 and 19 "9 7 inches did not find their way into India by way of 

 Persia, either directly or indirectly, but they may have been 

 modified, or carelessly applied; the sacred or royal cubits of 25*3 

 too were larger. I have taken from Mr. James Burgess's "Arch- 

 geological Survey of Western India " about 250 measurements of 

 from 2 feet to 100 feet • most from early Buddhist shrines, cells, 

 temples, and caves, as those of Elora, Anka, Kaladgi, Badamj, 

 Bhaja, Kuda, Gumli, Sana, Junagarh, Navalakha and Aurunga- 

 bad, say a.d. 200 to A.D. 1200. On tabulating these measure- 

 ments, I found a decided tendency to maxima and minima group- 

 ing of nearly five English feet ; giving, say, for maxima, the Nos. 

 60, 56, 51, 47, 42, 37, 32, 27, 22, 19, I6J-, 12J-, 7}, 4f, and 

 2*7. After trying various cubit lengths, say, of 17, 18, 19 and 

 20 inches, I found that a cubit of almost exactly 19 inches suited 

 best for a series of most likely cubits, giving, say, 38, 35J, 32J-, 

 30, 26J, 23J, 20J, 17 ? ^ i 2j 10 J, 8, 6, 4J, 2J, and 17, appar- 

 ently pointing to a series showing differences of 3 cubits, which 



