Al-Tabarani recorded it (6/188/5820). However, this Ibn Abi Sabrah is abandoned, as al-Haythami stated (6/252). Yet he was corroborated by the trustworthy Zayd ibn Abi Unaysah, whose reports were relied upon in the two Sahihs. However, the reports from him differ: Ubayd Allah ibn Amr narrated from him, from Abu Hazim, from Sahl, with the report. Al-Nasa'i recorded it in al-Kubra (no. 7299) through two chains from him, and its chain of transmission is authentic.

Abu Abd al-Rahim said: “Zayd narrated it to me,” but he did not mention Sahl, thus transmitting it as a mursal report. Al-Nasa'i recorded it as no. 7301, and its chain of transmission is likewise authentic, for Abu Abd al-Rahim—whose name was Khalid ibn Abi Yazid ibn Simak al-Harrani—was trustworthy and one of Muslim’s narrators. Likewise, the one who differed from him, Ubayd Allah ibn Amr al-Raqqi, was trustworthy and one of the narrators of both al-Bukhari and Muslim. In my view, his version is weightier for sound reasons:

First: al-Bukhari relied upon him, whereas he did not rely upon Abu Abd al-Rahim, as noted above.

Second: he was more closely associated with, and more retentive of, the hadiths of his shaykh Zayd ibn Abi Unaysah, as indicated by Ibn Hibban’s statement in his biographical entry in al-Thiqat (7/149): “He was a transmitter of Zayd ibn Abi Unaysah.”

Third: he has a corroborating narrator, namely Fulayh ibn Sulayman, as mentioned above.

The version recorded by al-Tabarani in al-Awsat, no. 652, through the route of Mu'allal ibn Nufayl al-Harrani, who said, “Ubayd Allah ibn Amr al-Raqqi narrated to us from Zayd with the report,” does not undermine this preference of mine. However, he transmitted it as a mursal report, omitting Sahl. I say: This does not detract from what I have stated, because Mu'allal is not well known, and no biographical notice has been written for him