GROWTH AND DECLINE OP CULTURE. 157 



brandies contain tlie calves^ according to age and sex^ tlien 

 tlie sheep in several subdivisions, tlie number of foxes killed, 

 tbe quantity of salt used, and, lastly, tlie particulars of the 

 cattle that have died. On other quipus is set down the pro- 

 duce of the herd in milk, cheese, wool, etc. Each heading is 

 indicated by a special colour or a differently twined knot. 



It was in the same way that in old times the army registers 

 were kept; on one cord the slingers were set down, on another 

 the spearmen, on a third those with clubs, etc., with their 

 oflBcers ; and thus also the accounts of battles were drawn up. 

 In each town were special functionaries, whose duty was to tie 

 and interpret the quipus ; they were called Quipucamayocuna, 

 or Knot-officers. Insufficient as this kind of writing was, the 

 official historians had attained, during the flourishing of the 

 kingdom of the Incas, to great facility in its interpretation. 

 Nevertheless, they were seldom able to read a quipu without 

 the aid of an oral commentary ; when one came from a distant 

 province, it was necessary to give notice with it whether it re- 

 ferred to census, tribute, war, and so forth. In order to in- 

 dicate matters belonging to their own immediate district, they 

 made at the beginning of the main cord certain signs only in- 

 telligible to themselves, and they also carefully kept the quipus 

 in their proper departments, so as not, for instance, to mistake 

 a tribute-cord for one relating to the census. By constant 

 practice, they so far perfected the system as to be able to re- 

 gister with their knots the most important events of the king- 

 dom, and to set down the laws and ordinances. In modern 

 times, all the attempts made to read the ancient quipus have 

 been in vain. The difficulty in deciphering them is very great, 

 since every knot indicates an idea, and a number of inter- 

 mediate notions are left out. But the principal impediment 

 is the want of the oral information as to their subject-matter, 

 which was needful even to the most learned decipherers. How- 

 ever, should we even succeed in finding the key to their inter- 

 pretation, the results would be of little value ; for what would 

 come to light would be mostly census-records of towns or pro- 

 vinces, taxation-lists, and accounts of the property of deceased 

 persons. There are still some Indians, in the southern pro- 



