Tom has a web page on this I'm trying to decipher:

http://www.rzuser.uni-heidelberg.de/~tdent/septenarius.html

It it he presents the deviations from JI fifths for his two
reconstructions of this temperament in terms of a new (to me) and
horrible notation for positive rational numbers, whereby p/q > 1 is
written +p:q and p/q < 1 is written -p:q. Could we PLEASE stick to the
standard mathematical notation everyone learned in grade school?

Now that I've vented about that (sorry Tom, but you hit my pet peeve
button) I'll give the deviations for "D=175", starting from C-G and
working the circle of fifths around to F-C: 392/393, 524/525, 350/351,
1, 1, 416/417, 1, 1, 1, 1, 440/441, 1

Now, a problem with this is that the prodcut of this deviations isn't
524288/531441, so octaves are slightly tempered. Presuming this isn't
intentional, at least one of these ratios is off, and in fact G#-D# is
given as 496/496. If instead we make that 4448/4455, we get untempered
fifths. In the vague hope that this is more or less what is intented,
here's a first go at what this temperament would be:

! sep.scl
Septanarius scale?
12
!
1568/1485
28/25
196/165
49/39
4/3
196/139
196/131
784/495
196/117
98/55
49/26
2