Reconstructions

Proto-MSA
*ḳny - to rear (Kogan 2015:561)
Leslau is probably correct in his identification of Proto-MSA *ḳny ‘to rear’ with PS *ḳny ‘to buy, to acquire, to possess’ (for which see HALOT 1111, CDG 437 and Kogan 2015:89), but the semantic shift is peculiar and has no immediate precedent elsewhere in Semitic. One wonders whether the meaning “to suckle” for Mhr. ḳənū and Jib. ḳéní may have developed by contamination with PS *ynḳ, which otherwise left no trace in MSA (cf. SEDNo. 83v). Tgr. ḳäna ‘to grow together’ (WTS 253) is too isolated to be taken as a reliable cognate in spite of the remarkable semantic proximity. (Kogan 2015:561)
Proto-MSA
*ḳry - to hide (Kogan 2015:562)
The origin of Proto-MSA *ḳry ‘to hide’ is uncertain. Contra Leslau, there is hardly any connection between this root and Arb. qry ‘to entertain a guest’ (Lane 2988). (Kogan 2015:562)
Proto-MSA
*ḳaṭmīm - fresh butter (Kogan 2015:562)
With Leslau 1947:9, Proto-MSA *ḳaṭmīm- ‘butter’ may be related to Arb. qaṭīmat- = ˀal-labanu l-mutaγayyiru l-ṭaˁmi (TA 33 289). (Kogan 2015:562)
Proto-MSA
*lV- - against, on (Kogan 2015:562)
As duly recognized by both Leslau and Johnstone, the MSA prepositions are semantically equivalent to PS *ˁalay rather than to PWS *li. The etymological *ˁ persists in the Soqotri forms with pronominal suffixes such as ˁəy ‘on him.’ (Kogan 2015:562)
Proto-MSA
*lˁb - to suck (Kogan 2015:562–563)
The origin of Proto-MSA *lˁb ‘to suck’ is unknown. Hardly any connection with either Hbr. lˁb ‘to make sport of’ (HALOT 532), Syr. lˁb ‘delectatus, gavisus est; avidus fuit’ (LSyr. 386), Arb. lˁb ‘to play, to joke (of children)’ (WKAS L 779) or luˁāb- ‘saliva, spittle’ (ibid. 809).
Proto-MSA
*lhy - to bleat (Kogan 2015:563)
The Proto-MSA verbal root *lhy ‘to bleat,’ probably onomatopoetic in its origin, has no direct parallel anywhere else in Semitic. (Kogan 2015:563)
Proto-MSA
*lhm/lḥm - to touch (Kogan 2015:563)
The origin of Proto-MSA *lhm/*lḥm ‘to touch’ is uncertain and it is even hard to say which of the two variants is primary from the etymological point of view. For *lḥm, cf. perhaps Syr. lḥm ‘aptus fuit; se applicavit; minatus est,’ (pa.) ‘conjunxit’, Arb. lḥm (VI) ‘to be joined, knit together’, Gez. ˀalḥama ‘to close, to glue’, Tgr. ˀalḥamä ‘to glue,’
Proto-MSA
*lkd - to patch leather (Kogan 2015:563)
Leslau is probably correct to connect Proto-MSA *lkd ‘to patch leather’ with Hbr. lkd ‘to catch animals by trapping’ (HALOT 530) and Arb. lkd ‘to cleave, to stick’ (Lane 2671), all going back to the presumably original meanings “to tie,” “to bind,” “to attach.”1466 Sab. mlkd ‘cistern’ (SD 82) likely belongs to the same root (with Biella 259: “something which traps the water brought by the canal”). (Kogan 2015:563)
Proto-MSA
*lsy - rain (Kogan 2011: 194)
Proto-MSA
*ma-lsay- - rain (Kogan 2015:543)
“One is tempted to connect Proto-MSA *ma-lsay- ‘rain’ with Akk. nalšu ‘dew’ (CAD N1 202, AHw. 724). The unusual sequence of two sonorants in the Akkadian word makes feasible the secondary origin of of n-, whereas the correspondence of Akk. (and PS) š to MSA s has good precedents elsewhere in the MSA vocabulary. This comparison is, however, quite tentative in any case.”