History of 馬 and 梅 I learned recently that two mora Sino-Japanese words using one character always end in /ki/, /ku/, /i/, /u/, /chi/, /tsu/, or /n/. However, I was also told that and are Sino-Japanese. What explains this difference from the trend? They are older words, but I also realized that, for instance, only has one mora in Chinese. (Although has two...) Did they just change over time?

As Snailboat mentioned in her comment, gogen-allguide.com has a good explanation for said etymologies.

Although, I should caution against analyzing Chinese moraically; Chinese is a syllabic language.

While {} and {} are etymologically borrowings from Chinese, they predate any of the three major classifications (ie, ), and are thus functionally nativized as indigenous Japanese lexemes.

The understanding is, at the time, Japanese had used a nasalized variant of /u/ (ie, /ũ/), which for /ũma/ and /ũme/ surfaced as [mma] and [mme]. Over time, this nasalization was lost, and thus, only the vowel quality /u/ surfaced, resulting in our modern /uma/ and /ume/.

(By the by, this nasalized /ũ/ also accounts for Chinese coda /ŋ/ ("ng") resulting in /u/, such as:

> /*kjiaŋ/ → /ki.ya.ũ/ → /kyou/