Maybe he used
sodium peroxide as an oxidizer - which it is. Then the
alkali might be the fuel, mixed with peroxide prior to ignition, since both would be solids.
A reaction looking like Na
2+Na
2O
2 2Na
2O
While such a reaction is exothermic, it makes a really poor pair of propellants, especially because it produces little gas to expand and because Na
2O melts and boils too easily, preventing to release the reaction heat, but few people would have been aware of that in 1927.
Any hydrocarbon would be a better fuel than an alkali, but it was possible to ignore it in 1927.
And, well, Na
2O
2 is a poor oxidizer as well... Even worse than H
2O
2.
One more difficulty: for a stable flame, solid propellants need a so-called "pressure exponent" below 1, which isn't very common and would need to be checked for Na
2O
2. This exponent tells how the speed of the flame front reacts to pressure. Knowing that the combustion chamber's throat evacuates mass with a pressure exponent of 1, this (which depends essentially on the oxidizer) tells if both combine harmoniously, if you see what I mean.
Perchlorates use to have a pressure exponent well below 1 hence are used, chlorates consistently make boom (unless the atmospheric counterpressure stabilizes the speed), and nitrates... sometimes.