50. § 49-50. 37 30 So always afa faffen, for this verb at the same time formally is a causative and as to its meaning (to teach) it belongs to the category, mentioned in 46. uàn b.) of the instrumental of the primitive's subject: Daç. 170 T A B Fat (she obtained an order of the king who was unaware [of what had happened before] to put to death this honest man); Mudr. I, p. 37 c T having got written the letter by Çakaṭadâsa); Panc. 51 a groei sangraraad (the cartwright let him bring home by friends), Kü- máras. 6, 52140191 gat-4 (he [Himavân] suffered his zenana to be entered by them, that is »he opened his zenana to them"), M. 8, 371 a saf: n (her the king should order to be devoured by dogs) ¹). aferen (after - In the passive construction these two types are likewise possible: 1. the primitive's subject turns nominative, the primitive's object remains accusative, as Mudr. V, p. 172 affanfan arr, the active form of which would be CHAT FRAT- am affenfadata, 2. the primitive's subject is instrumental, but the primitive's object turns nominative, as Mudr. I, p. 22 - 1) Panini gives a different rule about the construction of the causa- tives. In his sútra 1, 4, 52 he teaches that the primitive's subject is the karma of the causatives of a.) all intransitives, b.) the verbs of going (moving), c.) those of perceiving and knowing (fa), d.) those of feeding, e.) those of uttering voice, and the following rule declares »optionally also with entzufa and Erra [and their compounds, see Pat. I, p. 109, 1. 10]. With the other causatives, therefore, the primitive's subject is not considered an object (karma), accordingly not put in the accus., but in the instrumental, according to P. 1, 4, 55 compared with 2, 3, 18. Now, to these rules of Pâpini, which do not take account of the in- ternal difference existing by necessity between the two conceptions, but simply set up some outer marks, I have substituted the description expounded in the context. Mr. ANANDORAM BOROOAH has preceded me in this way. Moreover have tested Pâņini's rule in numerous instances, but found it deficient now and then even when paying due respect to the modifications made in it by the different vârttikas on our sûtras (1, 4, 52 sq.), whereas the same enquiry confirmed the exactness of the rule as it has been laid down in the context.
