MRC01, actually there is a bit of a paradox in the situation. ghellquist's reply is perfectly correct, but the directional pattern of an end-addressed, cylindrical microphone can never be perfect at high frequencies--while with a side-addressed microphone, the polar pattern in the horizontal plane can be audibly more consistent (though still not perfect) across the high-frequency part of the spectrum. So you generally have to choose between freedom of orientation and having the best available pattern at all frequencies.
In other words, fact #1 is that in an end-addressed cylindrical microphone, the pattern will be essentially the same in both the vertical and horizontal planes. That's just another way of saying that the microphone doesn't have a "top" or a "bottom" side--you can rotate it on its own axis while recording, and the sound you pick up won't change. But precisely because of the geometry which creates fact #1, fact #2 is that in either of the two planes (or any other plane that you choose), this pattern will vary at higher frequencies and at some angles of incidence. This is because sound at different angles of incidence will "see" a different shape than what front-arriving sound will "see", and will flow around or be reflected or diffracted differently.
Contrast this with a side-addressed cylindrical microphone: fact #1 will not be a fact any more (because in the vertical plane, the microphone has an entirely different shape than in the horizontal). But the situation for fact #2 changes, too, since no matter what the angle of incidence may be in the horizontal plane, the sound waves still come up against an identically-shaped and -sized side of the cylinder. As a result, the directional pattern in the horizontal plane can be somewhat better at high frequencies (of course that's just the housing--the capsule has its own acoustical properties, including its flat surface in the front and back). But you lose rotational symmetry completely, and the horizontal and vertical patterns can no longer be identical.
Does this make sense? It has some pretty interesting practical implications, I think. The cardioids I own which have the least peakiness off-axis at high frequencies are side-addressed, single-diaphragm cardioids. I like these microphones very much, as do many other classical musicians I've recorded with them. I'm at least as much of a gear slut as anyone else here is, yet as of autumn this year, I'll have owned these same microphones for 35 years--and if I absolutely had to sell all but two of my microphones, I would keep those two. (To be fair, I should explain that they're three-pattern microphones of a kind in which the omni setting is a true pressure transducer, unlike most other switchable-pattern microphones. Thus I would still have a considerable range of recording techniques available even if this gruesome choice were forced upon me.)
However, cardioids with a slightly brighter diffuse-field response seem to be more widely preferred in general. I don't always agree with this preference, but I do sometimes--for more distant miking in large, reverberant halls with smooth high-frequency absorption. Of my two favorite microphones of this second kind that come to mind, both are end-addressed--and that's no surprise when you consider the effects of the geometry.
--best regards