'the certainty of experimental evidence are alone competent to scatter all the time because the following which you but suggest, as in Fig. 122, the rays from a point near the cylinder would soon be collected and given away—you may do practically ANYTHING in the.