There is no fully statisfactory cognate for Proto-MSA *ŝVḫāp- ‘milk.’ Ugr. šḫp ‘colostrum, first milk’ would be directly acceptable both formally and semantically, but the hypothetical attestations of this lexeme in KTU 1.10 iii 25‒26 are highly doubtful. Syr. šḥāpā ‘colostrum’, on which the interpretation of the Ugaritic lexeme obviously depends, is also very sparsely attested and displays š instead of the expected *s, so that its similarity to the MSA forms may be purely accidental. More promising is Arb. šaḫb- ‘milk coming forth from the udder’ in spite of the irregular b. Of further interest can be Amh. šəffəta ‘globules of fat on the surface of milk’, End. šīf ‘clumps of butter coming out of the holes in the pot when the milk is being churned’. Almost certainly unrelated is Akk. šizbu ‘milk’, whose etymology is similarly unknown. (Kogan 2015:572)
“The origin of Proto-MSA *ṣaγlaf(-at)- ‘leaf’ is uncertain.1418 Leslau’s comparison with Hbr. sāˁīp ‘twig’ (HALOT 762) and Arb. saˁaf- ‘(leaves of) palm branches’ (Lane 1365) is, at the very best, highly tentative”. (Kogan 2015:543)
Also *ṣ̂aγṯ- ‘handfuľ. The origin of Proto-MSA *ṣ̂γṯ ‘to hold’ is unknown. (Kogan 2015:573)
“Proto-MSA *ŝiwāṭ- ‘fire’ has been correctly identified by Leslau with Arb. šawṭ- ‘heat’ (Lane 1619), Syr. sāṭ ‘combustus est’ (LSyr. 463). There is no trace of PS *ˀiš(-āt)- ‘fire’ in MSA.” (Kogan 2015:539)