68 TRANSACTIONS OF THE CANADIAN INSTITUTE. [Vo.. VI. 
become masculine or feminine. Modern Greek has preserved the neuter 
gender, and the signs of its disappearance in ancient Greek are not so 
numerous or clear as in Latin, but we shall see some of them, when we 
come to speak of the primary force of the ending a. But in Latin the 
tendency is very clear, and we have an exhaustive statement of the 
shifting in Appel’s De Genere Neutro Intereunte in Lingua Latina,* so 
that I shall content myself with citing a few of them here. Gladzus has 
an older form, gladium ; pulvis, m. and f., has an older form, pulver, n.; 
sanguen, the older form of sanguzs, and flumen are old neuters, as is also 
Venus. Biicheler, comparing volves with vzmum, speaks of otvos as the 
older, as if wine were naturally masculine. But in the time of Nero 
vinum had become vzxus, m.; for Petronius makes a peasant say vzuus 
mihe in cerebrum abit. Fatum has become fatws in Petronius. Jerome 
knows that cawbztwm is the classical form, but he uses czdztus, he tells us, 
to be generally understood. But the most striking proof of the extent 
of this process in popular Latin is the entire disappearance of the neuter 
gender in the romance languages. 
In Greek or Latin, adjectives agree in gender with the substantives 
with which they are in apposition, whether this apposition be attributive 
or predicative. I specify in this way, for it seems to me that the 
adjective in the predicate is much slower to enter on this agreement 
than the attributive adjective. In German the predicative adjective is 
invariable, and probably the similar lack of agreement in Greek in such 
examples as 03% dyadov zohvzotpaxty, point back to a time when this 
construction was the rule in Greek also. But with regard to their agree- 
ment in gender, adjectives in Greek and Latin fall into three classes: 
(1) Adjectives in most eommon use in Greek and Latin have three 
forms, one for each gender, ex. gr., dya6dés-y-0v or bonus-a-um. But 
interesting are those in Latin like ce/er-zs-e, where the distinction of the 
form celery as masculine from celerzs, feminine, seems to have been a 
novelty even in classical Latin, where Lucretius (IV., 160) has celer 
origo, and Virgil (Aen. VI., 685), uses alacris as masculine. In archaic 
Latin Neue notes the wildest confusion, indicating that no such 
distinction existed then. Parallel to this is what takes place with 
regard to certain common nouns in classical, as opposed to archaic 
Latin. In classical Latin Zzpa is the she-wolf, as opposed to /upus, and 
porcus has a feminine, forca. But Cato and Ennius wrote lupus femina 
and porcus femina. So Petronius forms apra from aper and Horace 
makes depus feminine. In this class, then, there is a manifest tendency 
to advance from a twofold to a threefold system of gender. (2) But the 
*Erlangen, 1883. 
EE a es 
