I wonder if your correspondents on Student's t have looked at p.666 of
Anders Hald's "A History of Mathematical Statistics From 1750 to 1930",
New York: Wiley, 1998. That seems to indicate the origin of the "t"
although the original sources should be checked.
On one of the matters raised later, about Russian statistics: Markov
discounted what he knew of the work of the English Biometric School,
which was brought to his attention by A.A. Chuprov. The correspondence
between Markov and Chuprov marks the coming together of probability
and statistics into mathematical statistics in prerevolutionary Russia.
Both M and C died in the early 1920's and Stalinism suppressed the
development of mathematical statistics until well after WW2 in the
Soviet Union. Few people wrote on statistical tests; Kolmogorov,
through his eminence was able to get away with it, but only just.
They did not, and still do not, have the common attitude to standard
tests that people outside have. Their thinking is either at the official
statistics end; or at he abstract probability end. Not what one would
call the inferential middle.
The French largely went their own way also, although people like Frechet
and Darmois tried to encourage "Anglo-Saxon" statistics. They tend not
to attribute anything to authors other than French (e.g. Laplace rather
than Gauss, for the distribution). Often they are right: as with the
Bienayme (rather than the Chebyshev) inequality.
Eugene Seneta