Intriguing, no? Wife/mother might well have been dead; or separated from her husband; or British already.
The document is dated 1898 and the 1901 census shows that the father is married. Now, why put that on the form if you don't mean it? So, wife/mother isn't dead then.
A chance find last week *may* have cleared up the matter: found on ancestry.com (in the library, you may recall, as it's free) a marriage between Herman and Adeline. Adding to the "possibility-quotient" was that Herman's parents had the same names as those cited in the naturalisation document.
And the marriage took place just before (well, a year before) the birth of the aforementioned son.
So, the PQ has just zoomed up higher. And there it will remain, hovering, until I can think of a way to confirm it. Good find, I thought!