A retina display is one that equals or surpasses the human eye in resolution - in other words where the resolution of the display is higher than the visual acuity of the eye. Specifically the fovia centralis which is the central region of the retina with the highest visual acuity, or sharpness.
The fovia centralis in the typical human eye can discern two lines as separate when they are more that 57 arcseconds apart, taking the eye as the center of the angle. That's 0.16 degrees apart. For instance, hold your hand at arms length and examine the hairs. You'll find that hairs that are very close together will appear to be one hair, but as you bring your hand closer you'll notice there are two or more hair where previously you thought there was only one.
Apple has determined (or decided...) that people hold their phones about 10" (25.4cm) from their eyes. At that distance the portion of the eye with highest visual acuity (fovia centralis) cannot discern two pixels that are 3 thousandths of an inch apart as separate pixels. The ipad is held about 15" away from your eyes, and at that distance your eye cannot discern two pixels that are 4 thousandths of an inch apart as separate pixels.
Therefore, at the ideal viewing distance, the resolution of these "retina" devices equals or exceeds the visual acuity of the portion of the human eye with the highest visual acuity - meaning that at those distances one cannot tell two adjacent pixels apart, and it appears to be one seamless image composed of continuous tones as one sees in nature, rather than a digital device composed of discrete lights.