Test tea

Oh, the making of words today is impossible, grotesque, line-and-arc sigils made of points of light that pretend to be bits of the world other than themselves.

I've been reading Constance Spry's 1961 Hostess today, which is, natürlich, entirely and unmitigatedly fabulous and like Liza Minnelli in her royal blue silk pyjamas at the Oscars last night made into a book. Though I'm perfectly certain Spry would have thought that look de trop. But on the other hand, the woman invented Coronation Chicken.

Hostess, which existed only in a draft when Spry died, is set in the aspic of wry wistfulness for how things were once. But she is Constance, and neither sentimental nor reactionary. Cream isn't served with tea as it once was, but no matter. We have milk, darlings.

She holds, nothing if not particular, that if the milk is added to the tea in the cup ...

... you can get 'a 'raw' taste rather difficult to describe; if you put it into the cup first the scalding tea poured on to it gives a mellower taste.

The milk-or-tea-first debate is a classic, burnished around the corners and comfy to sit on like a club fender. In 1919 a scientist called Muriel Bristol was working at the Rothamsted Experimental Station. The famous statistician R.A. Fisher, who also worked there, heard that Muriel claimed to be able to tell the difference between milk-first and tea-first cups.

Fisher tested Muriel's claim with a statistical significance test that came to be called Fisher's exact test, which has its own controversies, but was hugely influential and still bites the ass of research students today.

She passed.