Should be used as [彐(帚)], that is, the upper component of 帚.

u5F50 is analyzed as [彐(彑)], that is for snout, or the upper part of 彖.
u2B739 is analyzed as [彐(又)], that is for hand, or the lower part of 雪.

The evidence #1 support [彐(彑)] and [彐(又)]. ‘the snout, vulgar form [彐(彑)] ... and [彐(又)] is a different character’
The evidence #2 support [彐(又)]. ‘lower part of 彗 ...’
The evidence #3 support [彐(又)]. ‘variant of 又 (hand) ...’
The evidence #4 support [彐(彑)]. ‘variant of 彑 (snout) ...’
The evidence #5 support [彐(彑)] and [彐(又)]. ‘variant of 彑 (snout) ... used in 彗 ...’
So all the evidences are supporting [彐(彑)] and/or [彐(又)].

Note that, the upper part of 帚 becomes this shape, and we know [彐(帚)] is different from [彐(彑)] or [彐(又)].

⿲月⿳亠⿲纟⿱二口纟⿱⿲长马长心戈

▲ 留青日札, vol. 38

Update glyph to |⿰贝睿|.

The phonetic symbol of 𧸩 is the same as 濬 (璿, 䜜), is 睿 < 叡 < 㕡 *WEN.

Evidence #1 also shows a written form of |⿰贝睿|.

Change glyph to

02246

火 86.9.3

KC-10407

TS 13 · IDS ⿰火垕

Update glyph to match IDS and evidence.

Change glyph to

02272

火 86.12.5

KC-10072

TS 16 · IDS ⿱發火

Evidence 2 is unreliable.


▲ File:CADAL02096686 純常子枝語（四）.djvu

“坑” and “□” should be lam-alef ligature “لا” and lam [laːm] “ل”. So the character may pronounced “la”. From the letterforms to the phonetic notation, everything is incorrect.

All evidences support variant of 𫵮. So unify to 𫵮 (U+2BD6E).

Also, in SAT, 一切經音義: 從又持～啇, ～音瑟臻反, 古文作曹.
So according to SAT, it is used as 甡, so unify to 甡 (U+7521).

Now we know that the [彐(帚)] is actually [习].