The Jackson Cross Cylinder: How Your Axis Is Refined One Degree at a Time
"Which is better, one or two?"
Most people assume a refraction is a measurement, like standing on a scale. It is not. The automated instrument gives a starting point, and everything after that is a forced-choice experiment in which you are the detector. The tool that drives that experiment for astigmatism is a small lens called the Jackson cross cylinder, and understanding it explains both why the process takes as long as it does and why your answers matter so much.
What a cross cylinder is, optically
A Jackson cross cylinder is a single lens with equal and opposite powers in two meridians ninety degrees apart: plus in one, minus in the other. The usual clinical powers are plus or minus a quarter of a dioptre, and the lens is mounted in a flip handle so that the two positions can be presented alternately, with the handle held at a defined angle to the axis being tested.
Why flipping it is informative
The trick is what happens to the correction when you flip. In one position the cross cylinder adds a quarter dioptre of cylinder along the axis your glasses are meant to correct, and takes a quarter away at right angles. In the other position it does the reverse. Neither position is the "right" lens; both are slightly wrong, and they are wrong in opposite directions. Because the two errors are reflections of each other, a person comparing them is not being asked to judge absolute clarity, only which of two imperfect images is less imperfect. That is a much easier task, and it is why the answers stay meaningful even when the endpoint is close.
For the axis, the handle is aligned with the correcting axis and the lens is flipped. If neither position is preferred, the axis is already correct. If one is preferred, the axis is rotated a few degrees toward the position that the preferred flip implies, and the comparison repeats. Five-degree steps are typical, then one or two degrees as the endpoint approaches.
Then the power
With the axis settled, the cross cylinder is rotated so that its handle sits at forty-five degrees to the correcting axis and the flip is repeated. Now the comparison tests the amount of cylinder rather than its orientation, and the cylinder is increased or decreased in small steps until the two positions are equally acceptable. Because changing cylinder alone would also change the spherical equivalent, a compensating sphere change accompanies each step, which is why the numbers on the trial frame seem to move together.
The other checks that run alongside it
Two standard tests bracket the cylinder work. Fogging deliberately over-pluses the eye so that accommodation relaxes and the endpoint is not pushed myopic by effort. The red-green duochrome exploits longitudinal chromatic aberration: in a well-corrected eye the red target and the green target are meant to look similarly sharp, because the two wavelengths focus at slightly different distances. When the red side looks clearer, the eye tends to be relatively myopic; when the green side is clearer, relatively hyperopic. The separation between the two colours is roughly half a dioptre, which is why the test is used as a fine check rather than as a measurement.
What it means that the process is subjective
Because the endpoint is defined by your responses, its quality depends on the quality of the stimulus. A dry eye produces fluctuating optics between one flip and the next. A small pupil reduces the optical difference between the two positions. Fatigue, poor instruction and a hurry all degrade the result. This is why a refraction taken on a day when your eyes are streaming may not be the prescription you should order, and why a good optometrist will sometimes stop and treat the ocular surface before continuing.
Misconceptions worth correcting
- "The machine already measured my eyes, so the questions are a formality." An autorefractor is a starting estimate. Its output is refined precisely because it is not accurate enough on its own, particularly for cylinder axis.
- "One and two are random." They are two known, opposite errors. Answering on the basis of which is crisper is exactly the intended task.
- "More minus must be better because it looks darker." Extra minus can look subjectively sharper while over-correcting, which is what fogging and the duochrome are designed to catch.
- "A quarter dioptre makes no difference." At low cylinder values the difference between positions is small by design, but that is also why the axis can drift several degrees without complaint — and why small errors are tolerable at low powers and not at high ones.
When to raise it with your optometrist
Mention it if the two options seem impossible to tell apart, if your answer changes each time the same comparison is repeated, if vision fluctuates through the day, or if the final prescription feels worse than the trial frame did. Also mention previous surgery, keratoconus, or a history of irregular astigmatism, because those change how much refinement is meaningful and may call for corneal topography rather than more flips.
A quarter dioptre is small, and that is the point
Patients often ask why the process bothers with steps that seem too small to matter. The answer is that a quarter dioptre of cylinder at a given axis is not the same as a quarter dioptre of sphere. Cylinder error is directional: it blurs lines of one orientation more than lines at ninety degrees, and at the level of a quarter dioptre the effect on the endpoint is subtle but on axis it can be several degrees wide. The cross cylinder is designed to make that difference just detectable rather than to make it obvious, because the endpoint is defined as the point where the two alternatives stop being distinguishable.