Reconstructions

PES
*ṣ̂ew - salt (Kogan 2011: 239)
usually thought to be a Cushitism (but cf. Kogan 2006c: 271 for a tentative comparison with Ugr. ṣṣm ‘salt-works’, Hbr. ṣīṣ ‘salt’)
PES
*tkl - to fix, to fasten, to plant

Noeldeke 1886, 726 tentatively connected the Akkadian–Aramaic isogloss *tkl (considered to be a secondary derivation from PS *wkl) with PES *tkl “to fix, to fasten, to plant” . This comparison, implying that Ethiopian preserves the primary meaning “to fix, to be well-fixed, reliable” and undermining the derivation from *w-k-l, is discussed and rejected by Leslau in CDG 573.

PES
*tukʷān- - bedbug (SED II No. 122)
PES
*ṭbw - to suck (Kogan 2011: 228)
is derived from *ṭVb ‘breast’
PES
*ṭnn - to be asleep (Kogan 2015:574)

May be related to PES *ṭnn ‘to be asleep’ or Arb. ṭnn ‘to die’.

PES
*ṭirs­ - molar tooth (SED II No. 275)

Cf. *ṣ̂irš- - molar tooth. A variant root?

PES
*ṭlf - to snatch, to abduct (Kogan 2015:574)

An connection with Proto-MSA *ṭlf ‘to skim off’ is not unlikely. (Kogan 2015:574)

PES
*wˁy - to be hot, to burn (CDG 603; Kogan 2015: 464)
The origin of Proto-EthS *wˁy ‘to be hot, to burn’ is unknown (cf. DRS 577).
PES
*wγṭ - to swallow (CDG 611; Kogan 2015: 78)
PES
*wald-/*lVd- - son (Kogan 2015:458)
“The original meaning of Proto-EthS *wald- (and *lVd-) ‘son’ was probably “boy,” as is the case of its immediate parallels in Hebrew (yäläd, HALOT 412) and Arabic (walad-, Lane 2966). All these terms are derived from PS *wld ‘to bear’, widely attested both withing and outside EthS (SED I No. 80v). Lack of lexical distinction between the meanings “son” and “boy”/”child” — similar to bēn vs. yäläd in Hebrew — observable throughout EthS is atypical for Semitic languages and might be due to the substratum influence.”