Noting the weather at the opening of an entry as convention goes back to the Heian Period. Why! In fact, one particular Fujiwara Regent had people noting the weather for him. (That is to say, at least one did that got remarked upon in a study of courtly diaries.)

However, in recopying diaries, sometimes people left the weather out. Boo.

Having the weather data available makes it possible to compile all sorts of meaningless correlations. Do kitsune possessions increase with bad weather? Tanka writing with rainy days? Flower-viewing entries with wind?

Now I’m inclined to keep a Japanese diary once my Japanese improves enough. With weather, of course.

Also: what are Johnsonian words? :)

What a diary lover we are! In the Summer break, elementary school kids have to keep a drawing diary (e-nikki) for the whole summer.
The diary book has as many entries as the summer holidays. For 6 years, I endured!The first couple of days are OK but pretty soon entries begin to contain just weather and drawings of the sun and cicadas and other bugs. Missing entries of days start to appear.
I think I completely stopped writing in the first year.

I had to research dates for the obvious comparison to Ishikawa's "Romaji Diary," and indeed, it dates to 1909, a year before this Eibun Nikki.

@Leonardo Boiko,

"Johnsonian" refers to Samuel Johnson (1709–1784), the famous writer of "A Dictionary of the English Language" and a general man of letters.

Johnsonian words.

Thanks for covering the Johnson issue, you two.

MMS: I'd noticed it before but I didn't realize it went that far back! I checked my Tosa Nikki just now and that has dates but no weather, so I suppose the practice must have taken hold sometime around the turn of the millennium.

Anonymous: I understand that a lot of those e-nikkis are filled in all at once on the last evening before school comes back in...

Leonardo: I definitely recommend keeping a Japanese diary -- a handwritten one. It's one way to make sure you get at least a little practice writing Japanese every day (which means you can develop an actual, consistent, individual hand in Japanese, which you really can't do drilling rows of identical kanji).

Charles: You just reminded me of when I would tell students to write something in English and they'd just write Japanese in Romaji.

Basically, of all the things called 日記 (which could be just a list of documents filed by date), the 漢文日記 written on 具注暦 pretty much follow that pattern. I can't remember if 官職日記 like 外記日記 include weather, but that's a possibility. There might be some original copies without weather notations (I think the earliest manuscript, as opposed to copy, is from 御堂関白記), but that's almost the archetypical form.

Where it came from, I haven't been able to pin down. Tosa nikki doesn't read like many calendric nikki (neither does Murasaki Shikibu's, at least in part), and I'm not sure that they were really based on the same model, pace Ki no Tsurayuki's cross-dressing introduction.

Contemporaneous to Ki no Tsurayuki is Fujiwara no Tadahira's 貞信公記, but that mostly survives in a highly abbreviated form. I've been through it, but I don't remember if there's any weather notes.... By 小右記, definitely, they are there.

(What I should do is check if there's a Chinese precedent. But.)

The weather included tends to be related to precipitation (晴、陰／曇--which in practice means "looks like rain" or heavy cloud cover--, 雨降, 吹雪 and the like), and not to hot and cold except in rare exceptions. (And even then, the exceptions I can think of other than Teika's are in the body of the entry, as opposed to in the opening.) To this respect, it seems like weather is like noting wether the day is 仏滅 or the like--as some court procedures varied on whether it was fair or rainy, that might be the origin.

But that's just a personal theory.

Ah, my memory (and research) is faulty. I mis-recollected that Ishikawa wrote his diary in English, but it was actually written in Japanese in romaji. Mea Culpa. But your anecdote was amusing.

Didn't he write it in English as a kind of code? Or substitution cypher, technically. Worst one ever, too.

MMS, as ever I appreciate the info. I had never even considered that the weather-noting might have had a practical function.

There's about one passage of English in Romaji Diary: the rest (bar names of days of the week) is just Japanese in romaji. He sort of suggests he's writing in romaji so his wife can't read it, but -- yeah, worst substitution cypher ever.