▲ 東維子文集（明刊本）卷24 folio 8 《故翰林侍講學士金華先生墓誌銘》

On the first evidence, another text gives 溯, a semantic variant of 㴑:


▲ 景定建康志（文淵閣抄本）卷31 folio 37a

On the second evidence, another text gives 㴑:



▲ 雍正浙江通志（光緒刊本）卷239 folio 10a

Note that 雍正浙江通志（雍正刊本）卷239 folio 10a gives ⿰氵庠:



so the editor of 光緒刊本 revises ⿰氵庠 to 㴑 with intent.

⿰氵庠 might come from the cursive script form of 㴑:



▲ 字統网/㴑

Based on the evidences above, I suggest we mark ⿰氵庠 as a semantic variant of 㴑.