________________
50
ion" is identical with the phrase in the question preceding the programme. In a general manner, it can be said that davva-loga is closely connected with HAMG, whereas HAMG is absolutely absent in davvao. However, only $$21-24 employ HAMG in a uniform manner. - The above list indicates the area of conformity reaching up to the answers given in reply to the "repeated question" (viz. compounds in all cases except §§27 and 31). After this point, the differences amongst the different nikṣepas are so great that conformity of all the nine specimens or an appreciable part of them can nowhere be demonstrated. The programme is the centre of the nikṣepas, and the highest degree of conformity is found in the text immediately preceding or following it. The greater the distance from the centre, the greater the discrepancies.
It is, however, worthy of note that the extension of the compound (davva-loga) has been carried out rather systematically. Below we supply the evidence for all the davva-loga nikṣepas, giving in all cases the maximum length, i.e. the longest chain of compounds available: -
$20A
$21
$22
$23
924
$25
$26
$27
528
$29
$50
$31
$32
$33
THE CANONICAL NIKSEPA
152
Rayanappabhäpudhavi-aholoya-khetts-LOE
neraiya-davv'-AVICIYA-MARANE
neraiya-davv'-OHI-MARANE neraiya-davy-AINTIYA-MARANE
neraiya-davv'-EYANA
s' aiya-vīsasa-[davva-]BANDHE kala-MĀSĀ, attha-[davva-]MĀSĀ davva-PARAMAŅU
davva-SOE
uvagarana-davv'-OMOYARIYĀ
koha-kasaya-[bhāva-]VIUSSAGGE
davv'-INDIYĀ
kal'-OMAŅAM
PADI BANDIE
savva-mulaguna-PACCAKKHĀNE
4 members
3 members
3 members
3 members
3 members
3 members
2 members
2 members
2 members
3 members
3 members
2 members
2 members
1 member
3 members
No doubt, the formation of chains with more than two members is not found in all cases, and there also remains the problem of omitted members (square brackets). Even then we can say that the tendency to form compounds with three (and theoretically even more) members was very strong in the davva-loga area. These chains do not necessarily show an identical logical structure but, generally speaking, the tendency to form compounds over and above the "davva-loga" nucleus (and the partiality for the "repeated question") clearly demonstrate that davva-loga is basically different from davvao. Both types are ramificatory, but davva-loga presents a system of coexistent realities which may have considerable depth. By contrast, davvao has mutually exclusive statements which
