More helpfully.
Let us suppose that some time in 1905 (publication of the Rutherford theory of Radium decay/transmutation of that element), someone managed to convince a government to spend millions or billions on 'atomic power'.
Now. We know by that point, that radium gives off heat, that isotopes exist, and that some isotopes of some elements also release radiation.
We also know that most natural elements have both stable and unstable isotopes.
However, natural radioactive materials are exceedingly rare, or emit relatively little energy. Radium, a relatively common one apparently occurs at levels of grams per tonne of Uraninite (a Uranium ore of mostly UO2). Whereas Uranium and Thorium in a natural state aren't even particularly warm.
So. We can mine thousands of tonnes of Uranium ore to get kilo or so of Radium which could possibly boil minuscule amounts of water to run a steam engine that produced, what, a few Watts of electricity?
So natural Radium isn't the way to go.
Sooo...
You could try bombarding Lead with Oxygen atoms (or probably something a touch heavier), and separate out the resulting individual atoms of Radium.
You could try bombarding Molybdenum with protons to get *Technetium (OK, it hasn't been discovered yet. But the mere fact that it HASN'T means that it must have only radioactive isotopes. And the theoretical properties of 'ekamanganese' are relatively well understood, so being able to separate out the new element when it has been created is straightforward.
However, both these require particle accelerators - which haven't been invented yet. Well, that's OK, we know basically what's needed, in terms of theory - ions and electrons have been accelerated to relatively low energies (e.g. in Xray machines - the Xrays are produced by accelerated electrons).
However, it turns out that when OTL atom smashers were first built about 1930, they needed 600kV voltage. Which wasn't available anywhere. Fortunately Livermore got his hands on a recently published article on how to (theoretically) up high tension voltages.
So. If you want to go this route, you need to first develop the high voltage electronics a decade early. Doable, no doubt, but bloody expensive.
So. Once you've spent millions on buying pitchblende ore, monopolizing the world's supply of radium, and sending exploration parties out to find new radium mines around the world, and build mines in these remote locations; and spent more millions and millions on, firstly developing high voltage electronics, invented cycltrons (or other particle accelerators), built new power plants to power all this; THEN you can start running your transmutation machines to create artificial Radium and Technetium.
If you're lucky, you'll be getting microgram amounts.
OK. So this isn't the way to go. What else do we know. Well, how about fusion? That should be easy, right?
Well, iOTL, the masses of light elements were only measured sufficiently accurately in 1929.
Let's suppose that the Kaiser, or whoever is funding this says 'hey, i've got a few billion marks I have no use for, let's set up a lab to measure the masses of elements to unprecedented degrees. My 5 year old daughter insists that the sun must be fusing hydrogen, let's see if that's even possible.'
So they start hydrogen fusion work. Since that STILL doesn't produce net power today, with modern tech, you can guess how likely it is that it would work in the 30s.
=====
So, any foreseeable way forward at the time would be a VERY expensive dead end. In fact that amount of money wasted on that stuff would discredit pure research, and likely DELAY the discovery of fission. IMO.
-----
Note, too, that the whole 'spend billions on pure research' is a RESULT of the atomic bomb. Before that, the closest you got was things like the RN sending Cooke to Tahiti to observe the Transit of Venus, which is a much, much smaller scale project.