In practically making the test with a petrographical microscope remove the polarizer, analyzer and condensing lens,[36] and introduce below the section a small light-stop, to reduce in size the cone of incident light.[37] The adapter, holding this light-stop, should be adjustable so that the stop can be lowered sufficiently to produce the best results. The smaller the contrast between the indices of refraction the smaller the incident cone of light should be, as best results are obtained when this cone is little larger than twice _O′H_, Fig. 10, all the _O′H_ rays being then totally reflected. The more of the _HI_ rays that pass through, the brighter will be the _OD_ cone and the less sharp the contrast in illumination. A high power objective[38] should be used, as those with small aperture and great focal length do not give good results. Focus on the dividing plane and adjust until equal illumination is observed on both sides and the trace of the plane resembles a fine thread, the focal plane being at _MN_. Then _raise_ the _objective_ slightly, thus moving the focal plane to _ST_, when a _bright line_ or band will appear on the side of the _stronger refracting_ substance, the width of the line depending on the contrast between _n_ and _n′_, becoming narrower as _n_ and _n′_ approach each other in value. On raising the objective still further the line broadens and finally disappears. If the objective is lowered instead of raised the reverse phenomenon will take place, the bright line appearing on the side of the weaker refracting substance.[39] When the indices are nearly the same a bright line will appear on both sides and it is important then to make the cone of incident light as small as possible and to select the _brighter_ line. With doubly refracting minerals, each set of doubly refracted rays would suffer total reflection under somewhat different conditions and there would be no one point _B_ (as in isotropic minerals) where no bright lines would appear. However, if the cone of incident light is small enough in diameter, the test will be practically obtained as described. The Becke test can often be made with a medium power objective[40] without removing the polarizer and condensing lens, provided the condensing lens be lowered sufficiently. It may also prove convenient to turn the reflector a little sideways. The thinner the section the more distinct will be the phenomenon. The contact plane must be clear and not coated with opaque decomposition products. When the plane of contact _OO′_ deviates considerably from parallelism with the optic axis of the microscope a disturbance of the phenomenon may be expected and no satisfactory results be obtained. In general under these conditions the “bright line,” both on raising and lowering the objective, will remain on the side of the overlapping substance, without regard to the relative values of the indices.