Abstract In the examples we have considered so far, error introduced by numerical scheme has been O ( ∆t ). This can easily be improved as will show through two simple examples.

SOME time ago an instrument was described in this Journal intended for testing dark-adaptation in -normal and sub-normal patients. However, when it was tried to investigate gross disturbances (e.g., retinitis pigmentosa), the need was felt for a wider range of readings. When re-designing the apparatus provisions were also made to change the colour of the test-light if desired in order to make t...