H2 naturally occurring on its own for our purposes and even today can be relatively expensive. Where were they getting the hydrogen and how were they producing it? A normal SMR process using fossil fuels as a source?
The main method for industrial hydrogen making from 1904 was the "steam-iron" method. The actual method is older than 1904, but it was then that Howard Lane developed the first industrially-successful implementation (demonstrated at the St. Louis Exposition in 1904, where it was used to inflate balloons for the Gordon-Bennett balloon race (which became the Gordon Bennett Cup in ballooning in 1906).
Hydrogen is very common - our oceans are full of it. However, in the form of water, the hydrogen is bound to oxygen. Oxygen loves to bind to stuff! The trick is give the oxygen something it wants to bind to more than hydrogen. Iron is such a substance. Take hot iron, and blow hot steam over it, and the iron rips oxygen out of the water, producing a mix of steam and hydrogen. In terms of the physics/chemistry, it needs to be hot to overcome the energy barrier between the two states. Remove the water from the steam-hydrogen mix, and you have pure hydrogen.
To keep this going, you have to get that oxygen off the iron so you re-use it. This is done using "water gas", which is a mix of steam and carbon monoxide. The carbon monoxide pulls the oxygen off the iron oxide to make carbon monoxide and iron (this is the chemistry of making iron from iron ore (iron oxide)). The water gas is made by passing steam over hot coal or coke (i.e., carbon).
Lane's original hydrogen generator was improved, and in the 1930s, the two common systems were the Messerschmitt and the Bamag generators.
The advantage of having the hydrogen mixed with steam rather than some other gas is that water (i.e., steam) is easy to remove, in order to get pure hydrogen. The biggest problem with impurities in practice is usually hydrogen sulphide gas, if the coal/coke has too much sulphur. The are other methods to make hydrogen, but extracting hydrogen from the mixture is much harder. For example, coal gas (from the destructive distillation of coal) is usually about 50% hydrogen, with the rest being methane and carbon monoxide. It can be used as-is for balloons, since its density is about 0.7 times that of air, but this doesn't provide enough buoyancy for dirigibles like zeppelins.
In the 1940s, the steam-iron process was largely replaced by which skips the iron oxide step: just mix steam and a hydrocarbon gas (usually natural gas, which is mostly methane), with a nickel which acts as a catalyst. If the whole system is at about 1000C (again, must overcome that energy barrier), you get hydrogen and carbon monoxide and some leftover steam.
For a history of the steam-iron process, written by an insider who was involved in the construction of many of the early Lane plants, see