CHAP. XXIII. ARYAN HYPOTHESIS AND CONTROVERSY. 455 



sition it would be impossible to account for their structure 

 and composition, — as, for example, for the forms of the auxi- 

 liary verb ''to be," all evidently varieties of one common 

 type; while it is equally clear that no one of the six affords 

 the original form from which the others could have been 

 borrowed. So also in none of the six languages do we find 

 the elements of which these verbal and other forms could 

 have been composed ; they must have been handed down as 

 relics from a former period, they must have existed in some 

 antecedent language, which we know to have been the 

 Latin. 



But, in like manner, he goes on to show that Latin itself, 

 as Avell as Greek, Sanscrit, Zend (or Bactrian), Lithuanian, 

 old Sclavonic, Gothic, and Armenian are also eight varieties 

 of one common and more ancient t3-pc, and no one of them 

 could have been the original from which the others were 

 borrowed. They have all such an amount of mutual resem- 

 blance as to point to a more ancient language, the Aryan, 

 wdiich Avas to them what Latin was to the six Romance lan- 

 guages. The people who spoke this unknown parent S2:)eech, 

 of which so many other ancient tongues were offshoots, must 

 have migrated at a remote era to widely separated regions 

 of the Old "World, such as Northern Asia, Europe, and India 

 south of the Himalaya.* 



The soundness of some parts of this Aryan hypothesis has 

 lately been called in question by Mr. Crawfurd, on the ground 

 that the Hindoos, Persians, Turks, Scandinavians, and other 

 people referred to as having derived not only words but gram- 

 matical forms from an Aryan source, belong each of them to 

 a distinct race, and all these races have, it is said, preserved 

 their peculiar characters unaltered from the earliest dawn of 

 history and tradition. If, therefore, no appreciable change 



* Max Miiller, Comparative Mythology. Oxford Essays, 1856* 

 30 



