《漢俄合璧韻編》是十九世紀末俄羅斯漢學研究的重要里程碑。該詞典由巴拉第主筆、柏百福增補，前後費時十餘年，1888年於北京同文館出版。《韻編》廣泛收集口語與書面材料的字詞，書中記錄的漢字皆附有以俄文字母拼寫的漢語讀音。本論文即以俄文拼音為基本語料進行語音分析。《韻編》全書以俄文字母排序、以漢語的韻母分組。同一組內的漢字，有相同的聲母和韻母拼音，但沒有形式上的聲調標記；根據歸納結果，隱約依照「陰平、上、去、陽平」的聲調順序分組。本論文將《韻編》的俄文拼音調整為漢語音韻的格式之後，配合相關漢語音韻材料進行音位分析，試圖呈現《韻編》所記錄的語言；並且透過跨材料的比較，整理出本字典音系的語音特點和歷史意義。 　 　　一、平面音系分析。從時空背景和音韻特點來看，《韻編》所記錄的語言，應屬清末的官話。以下呈現漢語音位分析的結果。 聲母：　　p　pʰ　m　f　v　　t　tʰ　n　l　　k　kʰ　x　 　　　　　ts　tsʰ　s　　tʂ　tʂʰ　ʂ　ʐ　　ø 韻母：開　a　aj　aŋ　an　ɨw　ɨ　ɨŋ　ɨn　ə　ɨj　ɚ 　　　齊　jɨ　jɨŋ　jɨn　jə　jɨw　ja　jaj　jaŋ　jan　jaw 　　　合　wə　wɨ　wa　waj　waŋ　wan　wɨj　wɨŋ　wɨn 　　　撮　jwan　jwə　jwɨ　jwɨŋ　jwɨn 其中，有幾個韻母音位（如/jwə/）對應到兩組俄文拼音，只是因為聽感不同而有不同的拼音記錄，沒有系統上的對立。在聲韻配合方面，比較顯著的特色為：唇音不與合口呼、撮口呼韻母相配（除了合口呼單元音韻母）；軟齶音不與齊齒呼韻母相配；捲舌音不與撮口呼、齊齒呼韻母相配（除了齊齒呼單元音韻母）。 　 　　二、歷史比較。本論文將《韻編》音系與中古漢語和《中原音韻》比較，並與其他近現代外國對音材料比較。《韻編》與中古漢語的比較，有幾點特色。1.濁音聲母清化，平聲送氣、仄聲不送氣；2.知、莊、章系聲母多讀為捲舌音；3.不區分尖團音；4.沒有中古的入聲韻尾（塞音韻尾），而併入陰聲韻尾；5.陽聲韻尾則只有軟齶鼻音和齒齦鼻音兩類；6.[io]類韻母來自中古江、宕攝入聲韻。《韻編》與《中原音韻》的比較，比較明顯的是[m]韻尾讀為[n]韻尾，且《韻編》可以呼應若干條《中原音韻》的詞彙讀音註解。《韻編》與其他外國材料比較，不論俄、法、德、葡、英、美，經過音節比對，該時期官話在音韻系統上是相當一致的，沒有巨大的結構上的改變。不過就讀音變體而言，從清末至今，部分讀音漸趨弱化、消失，如韻母[io]弱化而與[üe]併合，但這類讀音的弱化或消失，並不影響整體音位系統。 / One of the most outstanding achievements of Russian Sinology during 19th century was “Chinese-Russian Dictionary” published in 1888. It was compiled by P. I. Kafarov and finished by P. S. Popov. A large number of Chinese characters, words and phrases were collected in this Dictionary, and Russian alphabets were supplementary to each written character for showing Chinese pronunciation. It is the Russian-Chinese transliteration that becomes our basic corpus. 　 　　In the Dictionary, the principle of listing order was by sequence of Russian alphabets, and principle of grouping was by Chinese finals (rhyme). In a group, every character had the same Russian transliteration. However, there was no marker of tone, which is important to distinguish meanings in Chinese language. To solve the problem, the editor or author indistinctly arranged the characters as Yin-ping (陰平 “dark level”), Shang (上 “rising”), Qu (去 “departing”) and Yang-ping (陽平 “light level”) four kinds of tones. After revising the Russian-Chinese transliteration to phonology form, phonetics analysis is possible. We would like to show the language recorded by this Dictionary, and summarize the feature of this language by comparing with different phonetic documents. 　 1. Phonetic analysis on the Dictionary. 　 　　Considering time, space and phonology feature of the Dictionary, we could claim that the language was Mandarin of late-Qing Dynasty. Here are the achievements of phonetic analysis. Initial: 20 initials grouping by place of articulation. 　　p　pʰ　m　f　v　　t　tʰ　n　l　　k　kʰ　x　 　　ts　tsʰ　s　　tʂ　tʂʰ　ʂ　ʐ　　ø Final: 35 finals grouping by jie-yin (介音 “medial; head of rhyme”), in order: kai-kou (開口“open mouth”) with no medial marker, qichi (齊齒“even teeth”) with medial marker /j/, hekou (合口“closed mouth”) with medial marker /w/ and cuokou (撮口“round mouth”) with medial marker /jw/. 　　a　aj　aŋ　an　ɨw　ɨ　ɨŋ　ɨn　ə　ɨj　ɚ 　　jɨ　jɨŋ　jɨn　jə　jɨw　ja　jaj　jaŋ　jan　jaw 　　wə　wɨ　wa　waj　waŋ　wan　wɨj　wɨŋ　wɨn 　　jwan　jwə　jwɨ　jwɨŋ　jwɨn A few of Chinese phonemes of rhymes like /jwə/ matched two Russian spellings. That was just the matter of listening difference and did not influence the phoneme system. Here are some remarkable features about the combination of initial and final. For example, labial consonants did not combine with “closed mouth” and “round mouth” rhyme, except monophthong of “closed mouth” rhyme; velar consonant did not combine with “even teeth” and “round mouth” rhyme, except monophthong of “even teeth” rhyme. 　 2. Historical comparison. 　 　　The Dictionary is compared with mid-ancient Chinese system and Mandarin document “Zhong-yuan Yin-yun”. Here are some remarkable features. First, total voiced initial consonants became devoiced, aspirate as level tone (平聲), while unaspirate as oblique tone (仄聲). Second, retroflex consonants were from Zhi (知), Zhuang (莊), Zhang (章) series in the mid-ancient Chinese. Third, there was no difference between sharp and rounded sounds (尖團音). Fourth, checked tone (entering tone) and its ending consonant were lost. Fifth, nasal endings were only [-n] and [-ŋ], while mid-ancient ending [-m] was lost. Sixth, rhyme [io] was from entering tone of Dang (宕), Jiang (江) rhyme group. Besides, we take some foreign documents in comparison. Our conclusion is that the structure of syllable and language system did not immensely change in that period. We can see that [io] rhyme became weak. Whether [io] rhyme existed or not did not influence the phoneme system.