Origins
Modern Japanese 十 read as じゅう comes from older Classical Japanese じふ, originally read as //d͡ʑipu//, in turn from Middle Chinese //d͡ʑiɪp̚//.
- The basic shift was for the //p// sound in //d͡ʑipu// to lenite (soften) into more like an //f// sound (specifically [[ɸ]], a bilabial, unlike the English [[f]] that is a labiodental). Then, the medial (mid-word) //f// sounds in Japanese underwent further lenition to disappear before everything but //a//. Lastly, the //iu// diphthong (two-vowel sound) flattened out into a //juː//, with the initial //j// here (like an English consontant-y "y") basically absorbed by the preceding front-of-the-mouth consonant //d͡ʑ//, resulting in modern じゅう.
//d͡ʑipu// → //d͡ʑifu// → //d͡ʑiu// → //d͡ʑuː//
じぷ → じふ → じう → じゅう
(The kana are not necessarily the actual spellings used, and are instead just meant to represent the pronunciations.)
Sound shift
There is a small number of kanji readings where the final //p// in Middle Chinese transforms oddly into a final つ. 十 is one such term, and 立 is another. This sound shift gave rise to the alternative on'yomi for 十 that we sometimes run into: じつ. This is a less-common reading, but it does crop up, seen in the 促音【そくおん】 or gemination that it causes.
This reading also appears in various terms, including but not limited to:
- 十戒【じっかい】 (the Ten Admonishments in Buddhism, also used for the Ten Commandments in the Abrahamic religions)
- 十界【じっかい】 (the Ten Realms in Buddhism)
- 十干【じっかん】 (the Ten Heavenly Stems in Chinese divination)
- 十脚目【じっきゃくもく】 (the order Decopoda, including shrimp, crabs, and the like)
- 十傑【じっけつ】 (the top ten in any listing of bests)
- 法華十講【ほっけじっこう】 (the Ten Lectures of the Lotus Sutra, a particular format for teaching the sutra)
- 十指【じっし】 (ten fingers; figuratively, a lot of people)
- 十手【じって】 (a kind of weapon used by constables in the Edo period)
Hypothesis: maybe there never was any reading じつ?
I cannot find any instances of 十 read as じつ, only where 十 is read as じっ with the gemination. One analysis of the //p// → つ sound shift could be that the Middle Chinese combination of the final //p// + initial following consonant was hard to hear as distinct consonants for the speakers of Old Japanese, and they may have interpreted combinations like //-ps-// or //-pt-// as simply geminate //-ss-// or //-tt-// instead.
Please comment if the above does not address your question, and I can edit to update.