漢語方言的音韻層次研院方興未艾，歷史音韻中異部接觸(出現於合韻、通假、聲古1[.....﹒)所反映的韻部之間的分合，具體而微體現了漢語白古即有音韻層次差異的一個面向。本文聚焦於音韻層次差異所體現的漢語音韻史意義，從合韻、聲訪1[、通假、方言異讀......的內容進行音韻層次的分析，探討經常發生接觸的之、脂、支三組韻部之間可能存在的時間、地域、或類型的差異。從文獻和方言材料的比較、分析，說明因為存在具體的層次差異，故韻部分合不能一刀切。段玉裁提出之、脂、支三分之後，蒸、真、耕雖然也對當三分，然而文獻資料和漢語方言卻有饒富意味的現象，說明事實並非分或合那麼單純。

In this work , 1 reconsider the hypothesis of a unitary sound system in Old Chinese , and argue that the Old Chinese sound system could well be a mixture of several dialectal accents. Scholars have been assuming a unitary sound system for Old Chinese , and diHerent classifications of rimes have been proposed. However, questions on riming across diHerent classes have been relegated to the periphery, and simply described 叫“ rhyme merger" 合韻phenomena. However, this phenomena may in fact reflect dialectal variations in Old Chinese. Were this indeed the case , at the time of Old Chinese , there should have been multiple (dialectal) sound systems rather than a unitary one , just as with the modern Chinese dialects Evidence from several sources supports this claim. In this work 1 focus on three groups of rhyme classes: [zhT, zhf, zheng] 之職蒸[zhT， zhf/zhì, zhen] 脂質真; and [zhT，丘， geng] 支錫耕A scrutiny of data from rhyming in literatu間， homonymous characters, annotations, and excavated inscrip tions shows that these three groups of rhyme classes were closely related in the eastern area of Northern China in ancient times and could freely rhyme with one another. As the close relationship among these three groups of rhyme classes is specific to a geographic area, it is unlikely accidental; the interrelation should reflect a common dialectal basis in the sound system of this particular geographic area. This claim also receives strong support from the Southern Min dialec紹remarkably ， the close relationship among these three groups of rime classes are phonetically transparent in Southern Min. That is , in the vernacular layer of the Southern Min dialects , the three rime classes , [zheng] 蒸， [zhen] 真， and [geng] 耕， all share the rhyme [-an] /[ -in]; [Zh1] 職， [Zh1/z hì] 質， and [X1] 錫share [-at]/[ -it]; [zh司之， [zl叮脂， and [zh T]支share [-ai]/[-i]. It has been proposed that one of the layers that composed the Min dialects originated from the eastern area of Northern China in ancient times. These thus suggest that the “ rhyme merger" , in the case of the three groups of rhyme classes alluded above , has a dialectal basis. If this analysis is correct, it is a strong piece of evidence for dialectal influence in the old Chinese , and the hypothesis of a unitary sound system should be reevaluated. This new position sheds light on some long-standing puzzles in the studies of historical phonology of Chinese , including the division among the rhymes [zh司之， [zh 司脂， and [zh司支， as well as the extraordinarily large number of rhyme classes for velar sounds (and the relatively few rhyme classes for bilabial sounds) in Old Chinese: they may have resulted from dialectal variations as well , as shown by the comparable phenomenon in Min dialects. AII these questions will receive intensive discussion in this work.