104 STUDIES IN ANIMAL LIFE. 



elements of which, they could have been composed. 

 If we find such forms as fai aime, we can explain 

 them by a mere reference to the radical means which 

 French has still at its command, and the same may 

 be said even of compounds like faimerai, i. e., je- 

 aimer-ai, I have to love, I shall love. But a change 

 from ye suis to tu es is inexplicable by the light of 

 French grammar. These forms could not have 

 grown, so to speak, on French soil, but must have 

 been handed down as relics from a former period 

 must have existed in some language antecedent to 

 any of the Eoman dialects. Now, fortunately, in 

 this case, we are not left to a mere inference, but as 

 we possess the Latin verb, we can prove how, by 

 phonetic corruption and by mistaken analogies, 

 every one of the six paradigms is but a national 

 metamorphosis of the Latin original. 



" Let us now look at another set of paradigms : 



Sanscrit. Jjj{^' Zend. Doric, sla O n ic Latin. Gothic. Armen. 



I am asmi esmi ahmi ('-/mi yesmo sum im em. 



Thou art 6si essi ahi kaoi yesi es is es. 



He is asti esti asti ka-ri yesto eat ist e. 



We (two) are . . 'svas esva yesva . . . siju 



You (two) are. . "sthas esta stho? to-rov yesta ... eijuts ... 



They (two) are. 'st&s (esti) sto? to-rov yesta ... 



We are 'sm&s esmi hmahi to-^er yesmo sumus eijum emq. 



You are Vtha este stha ka-re yeste estis sijup eq. 



They are stoti (esti) ht-nti JKT/ BOTnta sunt sind en. 



" From a careful consideration of these forms, we 

 ought to draw exactly the same conclusions ; first- 

 ly, that all are but varieties of one common type ; 

 secondly, that it is impossible to consider any of 

 them as the original from which the others have 



