Current reading: Alan Delgado's Have you forgotten yet? Between the Two World Wars. While this covers major national issues such as the General Strike, the Abdication and the Mosleyites. But it has its share of the small peculiarities of the era. 1924, for instance, saw a peak in a long-running craze for "Put and take", a gambling game using a six-sided top (one not so different from a dreidl and any number of similar devices, such as the "teetotum", going back centuries). That it had been a gambling device explains a memory from my childhood: how my grandmother reacted quite angrily when I got a yellow plastic put-and-take top as the toy from a Christmas cracker.
Another innovation of the 1920s was the publication of the first crossword puzzles in Britain (they had existed in the USA since the Liverpool-born editor Arthur Wynne introduced them to New York World in 1913). The first Sunday Express one in 1924 was a 7x7 non-cryptic effort that seems rather feeble by current standards; early puzzles also tended to go for densely interlocked grids of checked letters (in the style of the current "smallest, hardest crossword" format) but the familiar modern format and cryptic puzzles rapidly developed, notably via the poet and scholar Edward Powys Mathers (aka Torquemada).
I didn't realise until recently that there's a deal of philosophy behind crossword compilation. Torquemada is remembered as a great exponent and founder of the cryptic format, but many of his clues are imperfect by modern standards of clue-setting. Some were not even cryptic, in the crossword sense, just obscure general knowledge like completing literary quotations ("Chicken-skin, delicate, white, Painted by -- Vanloo").
A further problem relates to the logical basis, an idea codified by Torquemada's successor in 1939, Ximenes (Derrick Somerset Macnutt). Ximenes set the standard for the symmetric grid and the typical density of checked vs. unchecked letters (checked are the ones where words cross), as well as fairness of clues: no misdirection, guesswork, jumps of illogic, or breaching of the separation between the "definition" and "subsidiary indication" that comprise the clue. See Ximenean clueing and More thoughts on Ximenean clueing. Some puzzle setters such as the Guardian's Araucaria go against this convention; the rightness or wrongness of this is a matter of debate in crossword fandom.
If this kind of thing interests you, there's a lot more at Crosswords by Ximenes page (from which some of the above links come). The Fifteen squared weblog is a good stepping stone into the online crossword enthusiasts' circuit.