2 G. A. Grierson— On Irregular Causal Verbs [No. ], 
CENTRAL FAMILY. 
Wesr Centrat Group. 
Eastern Panjabi. Here in many verbs the causal suffix du is changed 
to lau or @l. H.g., / sikkh, causal, 4/ sikkha-l or Jf sikkh-l-Gu (also 
spelt sikkh-a-1 sikkh-l-@u). 
Gujarati. When the root contains any vowel except a or a, the 
causal suffix becomes d@-d;e.g., ./ bés, ‘set,’ causal, bes-d-d. If sucha 
root ends with a vowel or h, the causal suffix is vd-d, e.g., / lé, ‘take,’ 
causal, le-va-d. 
Rajputanz. Certain monosyllabic roots, ending in a vowel take 
y-Gv as the causal suffix; e.g., ,/ dé, ‘give,’ causal ,/ di-r-av. 
Hindi. Monosyllabic roots ending in a vowel change the causal suffix 
to la; e.g., »/ dé, causal 4/ di-l-a@. Soalso / sikh, causal 4/ stkh-l-G 
or 4/ sikh-G-l (dial.) (Old Braj ,/ sikh-a-r). 
These irregular forms do not occur in the Northern Group, nor in 
the Hastern Family (including Marathi). 
The origin of these forms has hitherto been considered obscure. 
Beames (Cp. Gr., ii, 81) says that they start ‘I know not whence, but 
- probably from a method in use in early Aryan speech, which has only 
been preserved by the classical language in a few instances.’ He then 
suggests that it is not impossible, and indeed it is highly probable, 
that the forms are connected by an interchange between the two semi- 
vowels 1 and v. Hoernle (Gd. Gr., 320) says, ‘the origin of these 
strange forms is very obscure.’ He rejects Beames’ hypothesis, but 
gives none of his own, beyond pointing out that certain Prakrit causals 
insert the letter d. 
Before proceeding to suggest an explanation, I must add another 
parallel form which has not been noticed by writers in this connexion. 
In the Marwari dialect of Rajputani, certain verbs insert an 7 before 
7j, the suffix which forms the passive. Thus the passive of lénd ‘to 
take’ is li-r-7jnd, and of dénd ‘ to give’ di-r-77nd. 
In Sanskrit, many verbs add consonants (usually nasals) to their 
roots to form the present stem. This, mdeed, is not peculiar to 
Sanskrit, but occurs throughout all Aryan languages. In Sanskrit, 
however, some of these consonantal root suffixes or infixes are formally 
recognized by grammarians, and are then called class suffixes (na, n@ 
and nu), and serve as the basis of arrangement of the verbal conju- 
gation. Examples are, (suffixes) kri-nd-ti, ‘he buys,’ sv-nd-tc, ‘he 
squeezes out,’ (infix) yu-na-k-tt, ‘he joins.’ When na is suffixed, the 
verb is not considered to belong to any special class, but is conjugated 
in the Ist or 6th class, thus vé-na-ti (Cl. 1), ‘he is anxious,’ pa-na-té 
