74 TRANSACTIONS OF THE CANADIAN INSTITUTE. [VoL. VI. 
menda, becoming an abstract noun, should become feminine, and why 
abstracts are generally feminine in Indo-European languages. Why 
should wzrtus, manhood, be feminine? Not surely through its meaning, 
as most enquirers seem to suppose, but probably by association with 
abstracts or collections ending in a, and these became feminine because 
of the common use of such abstracts or collections in a as epithets for 
females. How they came to be so used, is plain, I think, from the 
Homeric use of neuter plurals in a, ex. gr., in Iliad XI., 124: 
ypvad» AheFav0poto Osdsypevos, ayhad. Opa, 
or Iliad XIV., 238: 
Opa 0g tot Odcw xakodv Gpdvoy, 
where Spa is used as the epithet of a single object, with the meaning of 
an excellent gift. When a father wished to call his daughter “a good” 
or “a gift of God” he chose in preference to dya00v or Geddwpov the 
augmentative plural Agatha or Theodora. This augmentative form 
becoming the usual one for female epithets, was soon regarded as a 
feminine singular. Older female epithets, such as pyzye and Ovyazyp have 
no mark to distinguish gender and probably belong to the oldest stage 
of language, when genders were not distinguished, being formed after 
the analogy of zaz7jp. But with the adoption of @ as an ending for 
female epithets, the feminine gender was established on an equality with 
the masculine and neuter, and while in other languages we have a new 
distinction of genders according to sex, coming in to supersede the 
older, founded on civic status, in the Indo-European languages, the new 
distinction is naturally developed from the old, without obliterating it, 
though it gradually changes its character, till it seems as though 
distinction of sex were the idea on which distinction of gender was 
primarily based. 
. 
: 
| 
( 
