Note the use of the abbreviated form of koto (known as a gouryakugana) in the furigana for 「実」. The use of that kanji is itself unusual, in that 「実に」 is typically read 「じつに」 in modern Japanese. However, the kanji does have the alternate reading 「まこと」, so it is entirely permissible.
"You've lost a lot of weight, haven't you? (fig. 1) Well, your face's color isn't good — it's ghastly pale (fig. 2). Since when has it been this bad?", asked Yonehachi (fig. 3)
Fig. 1: 実にやせたねへ。
|Fig. 2: マア色のわるい|
The use katakana for the assertive sentence particle yo seems to be another established pattern in Umegoyomi. It also appears to be consistently written in a smaller size than the rest of the glyphs.
We also see the irregular use of periods once again — there appears to be no apparent grammatical reason for placing a period between 「は」 and 「真」, but there it is.
|Fig. 3: 何時分から|
The sentence-ending 「だへ」 seems to be historical kana orthography for 「だい」, which is still used informally as an interrogative sentence particle.
"Huh? It's been like this since after the 15th or 16th [of the month] (fig. 4). It's not that big of a deal, but I've been feeling depressed, which won't do" (fig. 5).
|Fig. 4: 主「ナニ|
|Fig. 5: 大造なこと|
The first word in fig. 5 was relatively easy to decipher in terms of kuzushiji by searching for the radicals, but the meaning of the word 「大造」, was only found after extensive searches online, which led me to this dictionary definition. It wasn't in the Koujien, but was in the Daijisen. Edit: from a comment below, it looks like it may actually an alternate spelling of the common modern Japanese word 「大層」. However, both words have similar meanings, so the end result is the same either way.
The kanji 「氣」 was a little easier to decipher thanks to my knowledge of the traditional version （旧字体） of the kanji 「気」. The kanji is part of a compound verb, 「気がふさぐ」, meaning "to feel depressed". What was really odd was the use of an unrelated kanji, 「閉」, to represent the end of the verb.
That the last kanji in fig. 5 is 「閉」 is only a guess. Anyone who has studied simplified Chinese characters knows that the radical 「門」 is simplified to 「门」. Many of the simplifications in Chinese come from traditional simplifications used in brush writing, and so I assumed that the kanji of interest would have a 「門」 radical and 3 more strokes below. The kanji that best matched this description was 「閉」, although I couldn't find a reading for it that matches this claim (therefore, other kanji might be possible matches too, so feel free to suggest any possibilities that come to mind).
I wasn't sure about the 「ならねへ」 at the end of fig. 5 either. I assumed that it was a dialectal version of 「ならない」, which is used in modern Japanese as a formal verb for expressing that something won't do. Any suggestions would be nice for this as well.
Note the use of an iteration mark in fig. 6. It looks more like the katakana iteration mark, 「ヽ」, than the hiragana iteration mark. However, since both serve the same purpose, it makes no real difference.
"But that's all fine — but how did you come to know [where I was]?(fig. 6) There are also many things I want to ask [you]." Tanjirou, moved to tears, looked pitiful (fig. 7).
Fig. 6: それはいゝ
The use of 「手めへ」 for 「手前」 was also very interesting, as it appears to be a precursor to the modern Japanese orthographical problem of mazegaki. Once again, the close, informal, relationship between Yonehachi and Tanjirou is established.
|Fig. 7: 聞てへことも|
Also of interest is that the word 「たんと」, which means "many" or "much", is possibly a Kansai-ben term. This term may have simply disappeared from Tokyo at some point later on in history.
The togaki contains a compound verb, 「なみだぐむ」, "to be moved to tears". You have to be careful here not to try to split up these verbs — always watch out for the continuous particle 「て」, because if it's absent, then you may have ended up splitting a verb when it shouldn't have been.
The final difficulty was the last kanji in the togaki, 「也」. The kuzushiji was impossible to decipher, but having heard the term 「哀れなり」 several times before, 「也」 came to mind fairly quickly. Looking at a kuzushiji database helped confirm that. Also, as you can see in fig. 1, it is the kanji from which the hiragana 「や」 is derived, and this character looks similar to 「や」.