It was also recorded by al-Sarrāj in his Musnad (fol. 95/1). This Ḥusayn is Ibn Bahrām al-Tamīmī, a trustworthy transmitter whose narrations are relied upon in the two Ṣaḥīḥ collections.
The report from Abū Hurayrah is preserved through six separate routes; I have documented all of them in my book Irwāʾ al-Ghalīl fī Takhrīj Aḥādīth Manār al-Sabīl, which I am presently authoring—may Allah facilitate its completion and publication. See no. 250 there.
I have chosen to comment on this particular chain because it alone contains the two additional clauses mentioned earlier. They define with precision what is meant by the word rakʿah (unit of prayer) in the various transmissions, namely: catching both the bowing (rukūʿ) and the first prostration together. Whoever misses the prostration has not caught the rakʿah, and whoever has not caught the rakʿah has not caught the prayer.
Benefits derived from the ḥadīth
It is therefore clear that the ḥadīth yields several important rulings:
First: it invalidates the view held by some schools that if the sun rises while a worshipper is in the second rakʿah of Fajr his prayer is nullified, and likewise that if the sun sets while he is in the final rakʿah of ʿAsr his prayer is void. This position is plainly false because it contradicts the explicit wording of the ḥadīth, as Imām al-Nawawī and others have stated. Nor may this ḥadīth be opposed by the reports that prohibit prayer at sunrise and sunset, for those reports are general whereas this one is specific, and in uṣūl al-fiqh the specific overrides the general.
One of the astonishing manifestations of blind partisanship to a madhhab over the ḥadīth is that someone will cite the ḥadīth to support his school in one issue yet contradict it in the very issue we are discussing; another will even declare the ḥadīth problematic for the same reason. To Allah we complain of the violations of the authentic Sunnah spawned by such sectarian zeal.
Al-Zaylaʿī said in Naṣb al-Rayah (1/229) after citing this ḥadīth of Abū Hurayrah and other reports to the same effect: