Most of the programming languages use // or # for a single line comment (see wiki). It seems to be that # is especially used for interpreted languages. According to this question the reason for that seems to be that one of the early shells (bourne shell) used '#' as a comment and made use of it (shebang).
Is there a logical reason why to choose # as a comment sign (e.g. symobolize crossing out by #)? And why do we use // as a comment sign in many compiled languages (especially in C as it seems to be one of the earliest compiled languages with that symbol)? Are there logical reasons for that? Why not use # instead of //, or // instead of #?
Is there a logical reason why to choose
#as a comment sign [in early shells]?
The Bourne shell tokenizer is quite simple. To add comment line support, a single character identifier was the simplest, and logical, choice.
The set of single characters you can choose from, if you wish to be compatible with both EBCDIC and ASCII (the two major character sets used at that time), is quite small:
! (logical not in bc)
#
% (modulo in bc)
@
^ (power in bc)
~ (used in paths)
Now, I've listed the ones used in bc, the calculator used in the same time period, not because they were a reason, but because you should understand the context of the Bourne shell developers and users. The bc notation did not arrive from out of thin air; the prevailing preferences influenced the choice, because the developers wanted the syntax to be intuitive, at least for themselves. The above bc notes are therefore useful in showing what kind of associations contemporary developers had with specific characters. I don't intend to imply that bc necessarily had an impact on Bourne shell -- but I do believe it did; that one of the reasons for developing the Bourne shell was to make using and automating tools like bc easier.
Effectively, only # and @ were "unused" characters available in both ASCII and EBCDIC; and it appears "hash" won over "at".
And why do we use
//as a comment sign in many compiled languages?
The // comment style is from BCPL. Many of the BCPL tokens and operators were already multiple characters long, and I suspect that at time the developers considered it better (for interoperability) to reuse an already used character for the comment line token, rather than introduce a completely new character.
I suspect that the // comment style has a historical background in margin notes; a double vertical line used to separate the actual content from notes or explanations being a clear visual separator to even those not familiar with the practice.
Why not use
#instead of//, or [vice versa]?
In both of the cases above, there is clear logic. However, that does not mean that these were the only logical choices available. These are just the ones that made the most sense to the developers at the time when the choice was made -- and I've tried to shed some light on the possible reasons, the context for the choices, above.
If these kinds of questions interest you, I recommend you find old math and science (physics in particular) books, and perhaps even reproductions of old notes. Best tools are intuitive, you see; and to find what was intuitive to someone, you need to find out the context they worked in. I am absolutely certain you can find interesting "reasons" -- things that made certain choices logical and intuitive to them, while to us they may seem odd -- by finding out the habits of the early developers and their colleagues and mentors.