his Sahīh (1), and the original report is in Sahīh al-Bukhārī (2).
This was his practice, peace and blessings be upon him, in all his prayers: making the earlier units longer than the later ones. He did this in the eclipse prayer and in the night prayer, when he prayed two long, long, long units, then two units shorter than the preceding two, then another two shorter than those before them, until he completed his prayer.
This does not conflict with his opening the night prayer with two brief units and instructing others to do so (3), for these two units serve as the opening to the night prayer. This prayer (4) is thus comparable to the sunnah prayer of dawn and other such prayers. The same applies to the two units he would sometimes pray after his witr, sometimes sitting and sometimes standing, despite his saying, “Make witr your last prayer at night” (5). Praying these two units does not conflict (6) with this
(1) No. 2738. Ibn Khuzayma (305, 944) and al-Tahāwī, in Sharh al-Ma‘ānī (1/183, 415) and Sharh al-Mushkil (4260), also transmitted this report through Mahbūb ibn al-Hasan, from Dāwūd ibn Abī Hind, from al-Sha‘bī, from Masrūq, from ‘Ā’isha. Ibn Abī Shayba (6774) transmitted an abridged version, and Ahmad (26042, 26282) and al-Bayhaqī (3/243) transmitted it through various chains from Dāwūd, from al-Sha‘bī, from ‘Ā’isha. Ibn Ma‘īn stated in his Tārīkh, as transmitted by al-Dūrī (3/485): “Whatever al-Sha‘bī narrated from ‘Ā’isha is mursal (disconnected).” Ibn Khuzayma identified this as the report’s defect, saying: “This is a gharīb (unusual) hadith. No one, to my knowledge, transmitted it with a connected chain except Mahbūb ibn al-Hasan. Dāwūd’s other transmitters narrated it as ‘from al-Sha‘bī, from ‘Ā’isha,’ with the sole exception of Mahbūb ibn al-Hasan.” This Mahbūb was rated weak by Abū Hātim and al-Nasā’ī. For the differences in its chains and routes of transmission, see al-Dāraqutnī’s al-‘Ilal (3620).
(2) No. 1090. Muslim (685) also transmitted it.
(3) Muslim transmitted the account of his practice from ‘Ā’isha (767), and his instruction from Abū Hurayra (768).
(4) That is, “this prayer.” The al-Risāla edition reads fahumā (“thus the two of them”); this, too, is an alteration by al-Fiqī.
(5) Recorded by al-Bukhārī (472) and Muslim (751), from ‘Abd Allāh ibn ‘Umar.
(6) This is the reading in all the manuscripts; the reference is to the prayer. Al-Fiqī amended it in his edition to lā tunāfiyān (“the two do not conflict”); the al-Risāla edition has the same reading.