Satapatha Brahmana Part V (SBE44), Julius Eggeling tr. [1900], at sacred-texts.com
13:3:5:11. Verily, there are deaths 3 connected with all the worlds; and were he not to offer oblations to them, Death would get hold of him in every world:
when he offers oblations to the Deaths 1, he wards off Death in every world.
13:3:5:22. Concerning this they say, 'If, in offering, he were to name them all, saying, "To such 2 (a death) hail! To such (a death) hail!" he would make that manifold death his enemy 3, and would give himself over to Death.' Only one oblation he offers to one of them, with, 'To Death, hail!' for there is indeed but one Death in yonder world, even Hunger 4: it is him he wards off in yonder world.
13:3:5:33. A second oblation he makes with, 'To Brahman-slaying, hail!' for, doubtless, a murder other than the slaying of a Brahman is no murder; but that--to wit, the slaying of a Brahman--is manifestly murder: he thus manifestly wards off Death 5.
13:3:5:44. Mundibha Audanya 1 it was who discovered this atonement for the slaying of a Brahman; and when one offers the oblation to the Brahmahatyâ he prepares a remedy for the slayer of a Brahman by satisfying Death himself with an oblation, and making a protection 2 for him (the slayer). At whosoever's Asvamedha, therefore, this oblation is offered, even if in after-times 3 any one in his family kills a Brahman, he thereby prepares a remedy (expiation) for him.
339:3 That is, according to Sâyana, on Taitt. Br. III, 9, 15, 1, causes of death, such as diseases, &c.
340:1 The oblations referred to in this Brâhmana (§§ 1-4) occur towards the end of the second set of 'aranyeऽnûkya' oblations mentioned above, p. 336, note 2, where the formulas are given. According to Taitt. Br., l.c., however, these final oblations are to be performed--like that to Gumbaka (Varuna)--at the time of the purificatory bath, which, indeed, may also be intended by our Brâhmana, though Kâtyâyana and Mahîdhara seem to offer no indications to that effect. It is clear that these final oblations must have formed the subject of considerable discussion among the early ritualists.
340:2 That is, according to Sâyana (Taitt. Br.), 'To death in the shape of disease, to death in the shape of poverty, &c.' Harisvâmin, on our passage, has merely, 'Amushmai pitrilokâya mrityave'--'To death (in the shape of) the world of the Fathers,' which is not very clear.
340:3 Or, perhaps, he would make himself many a death-enemy (bahum mrityum amitram kurvîta), the two nouns being taken as in apposition to each other; cf. p. 146, note 1.
340:4 See X, 6, 5, 1.
340:5 Or, he thus wards off what is manifestly Death (Death in person).
341:1 That is, according to Harisvâmin, the son of Udanya (Odana. St. Petersb. Dict.), Taitt. Br. III, 9, 15, 3, has Mundibha Audanyava (i.e. the son of Udanyu, Sây.) instead. The Taitt. Br., besides, makes the crime to be expiated here to be, not 'brahmahatyâ,' but 'bhrûnahatyâ,' the killing of an embryo. Sâyana, however, there allows to 'bhrûna' optionally its later meaning of 'a Brâhmana versed in the three Vedas and the sacrificial art (kalpa),' and the Taitt. Br. itself, at all events, takes this oblation to 'bhrûnahatyâ' to atone likewise for the slaying of a Brâhmana.
341:2 Harisvâmin explains 'paripânam' by 'parisishtam vânantam pânam' (?); whilst Sâyana, in Taitt. Br., takes it in the sense of 'sarvatah pâtram,' i.e. having made the Sacrificer 'a thoroughly worthy person.'
341:3 Harisvâmin here unwarrantably takes 'aparîshû' in the sense of 'in past times.'