In conclusion it seems desirable to make a few observations of a more general character about the method of worship described here.
The first conclusion may be that the temple worship described in the handbooks of Atri, Bhrgu, Kāśyapa and Marici is essentially based on the domestic worship found in the* Vaikhānasagṛhyasūtra*. There are many additions, and in some cases the order of the elements has been changed, but these are for the main part due to the more elaborate character of the worship in the temple and not to fundamental changes in outlook. In the worship proper, the pūjā directed to Vişņu (57ff.), the number of attendances has been increased to a total of thirty-two, but one is still able to recognize the original character of this pūjā: the invocation and reception of a divine guest and his entertainment. (For some further observations on the original character of the pūjā, one may be referred to Fausta Nowotny, in Indo-Iranian Journal, I, 1957, p. 110f.)
The conservative character of the Vaikhānasa method of worship can also be observed by considering the evidence of some “negative” features. Thus it was seen in group C (24ff.), that the personal preparation of the priest in which he identifies himself with Visņu in order to do the divine work, is, in the handbooks, of a rather restricted but original nature. The younger commentaries and monographs devote a much greater place to the practices of nyasa, syllable mysticism, etc., in accordance with Pāñcarātra and Śaiva systems.
A second “negative” feature in Vaikhānasa pūjā is the absence of any sign of the manasapūjā (mental worship), also called antaryāga- which is emphasized so often by the authorities on pūjā, e.g. by the Śrī Vaisnavas (Rangachari, The Śrī Vaishnava Brāhmaṇs, Madras, 1931, pp. 141 ff.), the Pāñcarātrins (e.g.* Lakṣmī Tantra*, ch. 36), the Śaivas (e.g.* Linga Purāṇa* 2, 23f.), and the Tantrists (e.g. Avalon, Principles of Tantra, pp. 1144ff.). True, the Vaikhānasa *arcaka *should meditate on the deity; and the handbooks of Bhṛgu (BKh and BK) even mention the worship of God in the heart (see 0.1), but this mental worship with attendances etc. just as in actual worship is not emphasized at all by the older Vaikhānasa sources. Of the younger monographs, AN devotes about one page to it (pp. 54f.).
An important difference with the *pūjā *of Tantric character edited and commented upon by F. Nowotny, op. cit., is that God is not thought to come out of the heart of the worshipper and to enter again into it after the worship has ended. The Vaikhānasa authorities only mention the invocation of God’s presence out of the immovable image into the movable image (by means of an intermediate stage, see 55), and at the end of the day again the dismissal of the divine presence into the immovable image. There is certainly a connection between this method of invocation and the absence of emphasis on mental worship.
A very original feature of the Vaikhänasa method is the nature of the formulas used. The number and importance of the mantras taken from the *Yajurveda *and the *Rgveda *is very great with respect to other *püja *systems. The short utterances and syllables (bijaksara) so well known from Tantric and other descriptions of pujä, are nearly entirely absent (but they have been introduced in great numbers by the younger sources such as KA in group C). But the most remarkable fact is that many of these mantras used by the Vaikhanasas are totally unknown anywhere else. In this survey there are 38 of them, spread over the whole ceremony (esp. in 24ff., 33ff., 38ff., 56ff. and the Appendix) but still many more destined for other uses than daily worship exist. They have been collected by the Vaikhanasas and edited under the name *Vaikhanasamantraprašna *(VMP, see Introduction), together with the Vedic mantras used by them.
A few words must be said here about the nature of this *Mantraprasna *or Mantrasamhită. It consists of eight chapters, the first four of which contain the mantras necessary for domestic worship; there it is nearly completely in accordance with the Grhyasutra. The last four chapters, called *daivikacatustayam *in the edition, are meant to accompany temple worship as described by a handbook. The question which handbook cannot be answered by a comparison with the respective descriptions of the daily worship as above. In fact it appears that the VMP sometimes shows an order of mantras which is found in none of the handbooks (40.3), sometimes gives formulas for attendances found at the place in question only in the younger sources (e.g. 36.3, 79.3, 79.9) or even formulas not found in the sources at all (40.3). In the great digressions found in 36 and 79 VMP tends to agree with those sources which have these digressions; they must be characterized as later additions. This all leads to the conclusion that the VMP cannot be considered as old, in any case in its redaction not as old as the oldest handbooks; this notwithstanding the fact that it has been provided throughout with accentuation in the Vedic method. Of course it contains materials of a respectable age, but the redaction of the VMP as known nowadays might be fairly recent. One has also to reckon here with additions made in the various manu scripts (see, e.g., 36.2), a fact which may also have been of some influence in the handbooks themselves, where from time to time verses have been put between brackets by the editors. It is possible that the redaction of the last four chapters of VMP took place much later than that of the first four destined for domestic worship. A thorough investigation of the whole Mantrapraśna will be necessary before a more reliable characterization can be given.
For similar reasons, it is difficult to try to place the handbooks themselves in chronological order: to arrive at reliable data the whole of these handbooks will have to be taken into account. Yet something may be said here with a fair degree of certainty. The texts belonging to the Bhrgu Samhita (BKh, BK, BY and BP) will have to be considered as secondary with respect to the other texts (A, K and M). BP instantly gives signs of its lateness by its number of additions and the younger and more systematic character of its treatment. But also BKh and BK show such additions with respect to A, K and M. Thus for example, they give a much more elaborate description of the bathing (36.2ff., 39.1, 40.1 and 2, 43); they add new elements in the pajd (especially BK is very detailed about the royal honours in 79.2ff.). In 47, they conceive the images of some minor deities near the image of Visnu as really present; BK presupposes a special halibera to be used for circumambulations around the temple (0.2.1, 0.3; cf. 87); in 54.1, only BK mentions the deities of the syllable a. BKh presupposes the matrkanyasa, *vyapaka *etc. in 40.1.
The signs of BY’s lateness are of a different character, but they exist none the less. This text gives an interesting addition about meditation on Vişnu’s weapons in 28.1. An argument of a more radical nature is, how ever, that in many places BY shows a striking dependence on the version given by M (e.g., in 40, 44, 68, 73.2). This dependence often even takes the character of similarity in wordings (a fact which of course cannot emerge from this survey which does not cite the sources directly). This similarity is clearly visible notwithstanding the fact that M is written in prose and BY in šlokas.
Because the handbook of Atri is written in slokas, it is possible to make a more precise investigation into the possible similarities between A and the texts ascribed to Bhrgu, which are all in slokas. Such a comparison reveals eloquent facts. BKh, ch. 20, has no less than 82 padas (one fourth ślokas) in common with A, ch. 40; and BK, ch. 9, has 39 pådas in common with A, ch. 40. The identity between BK, ch. 9 and BKh, ch. 20, amounts to 51 pādas. BY and BP, however, show a relatively great degree of individuality with respect to BK, BKh, A and each other. Very often the texts follow the device of changing one word in a pada or śloka (often only a small word like ca, tu, or a change of indicative into optative etc.; such padas have been treated as identical with each other in the figures given above).
In all the cases where identity with A exists it is, in my opinion, most probable that A represents the original; in the case of the identity of BK and BKh, the latter might contain the original, because it is sometimes cited by name by BK. The chapter on daily worship can lead to the assumption that A is the oldest of the handbooks (see its treatment of group Ci in a few places K gives the impression of paraphrasing it), but nothing certain can be said after an examination of only this limited subject.
In general, the Vaikhanasa method of performing devout temple worship in the description of the older handbooks can be characterized as elaborate, yet original and conservative; a real influx of new traditions, together with changes in the system, began in the descriptions of the school of Bhrgu, and continued in the monographs of the other schools (KA, KAV, AN). These new traditions show the influence of the Pan carātra and Śrīvaisnava schools, and of ideas fully developed in a Saiva and Šakta milieu; but at the same time the old traditions were maintained, as is shown by the fact that KA and KAV introduce a few other formulas from the Veda and in the order of the pūjā sometimes again rely on the description found in the Grhyasūtra.